Mathematical Modelling of Drug Delivery to Solid Tumour · Solid Tumour by Wenbo ZHAN A ......
Transcript of Mathematical Modelling of Drug Delivery to Solid Tumour · Solid Tumour by Wenbo ZHAN A ......
Mathematical Modelling of Drug Delivery to
Solid Tumour
by
Wenbo ZHAN
A Thesis submitted in fulfilment of the requirements for the degree of
Doctor of Philosophy and Diploma of Imperial College London
Department of Chemical Engineering
Imperial College London
September 2014
3
Abstract
Effective delivery of therapeutic agents to tumour cells is essential to the success of most cancer
treatment therapies except for surgery. The transport of drug in solid tumour involves multiple
biophysical and biochemical processes which are strongly dependent on the physicochemical
properties of the drug and biological properties of the tumour. Owing to the complexities
involved, mathematical models are playing an increasingly important role in identifying the
factors leading to inadequate drug delivery to tumours. In this study, a computational model is
developed which incorporates real tumour geometry reconstructed from magnetic resonance
images, drug transport through the tumour vasculature and interstitial space, as well as drug
uptake by tumour cells. The effectiveness of anticancer therapy is evaluated based on the
percentage of survival tumour cells by directly solving the corresponding pharmacodynamics
equation using predicted intracellular drug concentration.
Computational simulations have been performed for the delivery of doxorubicin through various
delivery modes, including bolus injection and continuous infusion of doxorubicin in free form,
and thermo-sensitive liposome mediated doxorubicin delivery activated by high intensity focused
ultrasound. Predicted results show that continuous infusion is far more effective than bolus
injection in maintaining high levels of intracellular drug concentration, thereby increasing drug
uptake by tumour cells. Moreover, multiple-administration is found to be more effective in
improving the cytotoxic effect of drug compared to a single administration.
The effect of heterogeneous distribution of microvasculature on drug transport in a realistic
model of liver tumour is investigated, and the results indicate that although tumour interstitial
fluid pressure is almost uniform, drug concentration is sensitive to the heterogeneous distribution
of microvasculature within a tumour. Results from three prostate tumours of different sizes
suggest a nonlinear relationship between transvascular transport of anticancer drugs and tumour
size.
Numerical simulations of thermo-sensitive liposome-mediated drug delivery coupled with high
intensity focused ultrasound heating demonstrate the potential advantage of this novel drug
delivery system for localised treatment while minimising drug concentration in normal tissue.
4
Acknowledgement
This thesis represents not only the result of my hard work over the last four years, but also
invaluable supports from sincere and kind people in my life.
First and foremost, I am grateful to my supervisor Professor Xiao Yun Xu. She has been
professionally and personally supportive since the day I applied for a Ph.D. position at Imperial
College. My Ph.D. experience would not have been as stimulating and productive without her
contributions of ideas and time. I appreciate her motivational enthusiasm and great patience to
guide me through the tough time in the Ph.D. pursuit.
Many people have helped and taught me immensely during this project. Thank you to Professor
Wladyslaw Gedroyc for kindly providing MR images and answering all my questions. I would
like to acknowledge Professor Simon Thom who gave me many suggestions from clinical
aspects to help improve my work. The time of working with Professor Nigel Wood is
unforgettable because of his great enthusiasm and appreciable willingness. I also want to express
my special thanks to Dr. Xin Zhang for sharing knowledge of thermo-sensitive liposome and
providing useful suggestions based on his experiments. Generous help was also given by Dr.
Cong Liu, who helped me a lot in deepening understanding of drug transport processes.
Our group has become a source of friendship, kind help and collaboration. My time at the college
could not have been as colourful or joyful without any of these. Thanks to Drs. Ryo Torii, Zhou
Cheng and Chrysa Anna Kousera, my fellow colleagues Afet Mehmet, Zhongjie Wang,
Harkamaljot Kandail, Shelly Singh, Oluwatoyin Fatona, Ana Crispin Corzo, Boram Gu, Claudia
Menichini and Andris Piebalgs. I wish you all the best.
My time in the U.K. was made enjoyable and cheerful because of many friends. I cherish the
time spent with friends in London, trips to other UK cities and abroad. Many thanks to my
friends for their long-lasting friendship over years. Da Gao and Wenbo Dong’s visits were
surprising gifts to me. Heartfelt thanks to Yijiu Jiang for his fraternal supports in the past years.
Last but not the least, I am deeply indebted to my parents for their love and encouragement under
any circumstances. They raised me with a love of science and supported me in all reasonable
pursuits. It is my great honour and happiness to be a member of the family. Thank you.
5
Declaration
I hereby declare that the work represented in this thesis was originated entirely from me.
Information derived from the published and unpublished work of others has been acknowledged
in the text and references are given in the bibliography.
The copyright of this thesis rests with the author and is made available under a Creative
Commons Attribution-Non Commercial-No Derivatives licence. Researchers are free to copy,
distribute or transmit the thesis on the condition that they attribute it, that they do not use it for
commercial purposes and that they do not alter, transform or build upon it. For any reuse or
distribution, researchers must make clear to others the licence terms of this work.
Wenbo ZHAN
Imperial College London, September 2014
6
Contents
1. Introduction ........................................................................................................................... 17
1.1. Background .................................................................................................................... 18
1.1.1. Cancer ..................................................................................................................... 18
1.1.2. Anticancer Therapy ................................................................................................. 19
1.1.3. Thermo-Sensitive Liposome-Mediated Drug Delivery .......................................... 22
1.2. Objectives of the Study .................................................................................................. 23
1.3. Structure of This Thesis ................................................................................................. 24
2. Literature Review ................................................................................................................. 25
2.1. Characteristics of Solid Tumour .................................................................................... 26
2.1.1. Tumour Vasculature................................................................................................ 26
2.1.2. Blood Flow Rate ..................................................................................................... 28
2.1.3. Lymphatic Drainage................................................................................................ 29
2.1.4. Interstitial Fluid Pressure ........................................................................................ 30
2.2. Modes of Drug Delivery to Solid Tumour ..................................................................... 30
2.2.1. Traditional Administration...................................................................................... 31
2.2.2. Nanoparticle-Mediated Delivery ............................................................................ 32
2.2.3. Polymer Implant...................................................................................................... 34
2.3. Drug Transport in Solid Tumour .................................................................................... 34
2.3.1. Transport within Blood Vessel ............................................................................... 35
2.3.2. Transport through Blood Vessels............................................................................ 36
2.3.3. Transport in Interstitial Space ................................................................................. 37
2.3.4. Transport to Lymphatic System .............................................................................. 39
2.3.5. Transport through Cell Membrane.......................................................................... 39
2.3.6. Cell Killing.............................................................................................................. 40
2.4. Summary ........................................................................................................................ 41
3. Mathematical Models and Numerical Procedures............................................................. 43
3.1. Mathematical Models ..................................................................................................... 44
7
3.1.1. Interstitial Fluid Flow ............................................................................................. 45
3.1.2. Drug Transport ........................................................................................................ 47
3.1.3. Thermo-Sensitive Liposome-Mediated Drug Transport ......................................... 49
3.1.4. Pharmacodynamics Model ...................................................................................... 51
3.1.5. Heat Transfer under High Intensity Focused Ultrasound (HIFU) Heating ............ 51
3.2. Model Parameters ........................................................................................................... 53
3.2.1. Tissue Related Transport Parameters ...................................................................... 54
3.2.2. Drug Related Transport Parameters ........................................................................ 54
3.3. Implementations of Mathematical Models ........................................................................... 59
3.3.1. ANSYS Fluent ........................................................................................................ 59
3.3.2. User Defined Routines for Drug Transport ............................................................ 60
3.3.3. User Defined Routines for Heat Transfer ............................................................... 60
3.3.4. Boundary Conditions .............................................................................................. 60
3.4. Summary ........................................................................................................................ 60
4. Non-Encapsulated Drug Delivery to Solid Tumour .......................................................... 65
4.1. Model Description .......................................................................................................... 66
4.1.1. Model Geometry ..................................................................................................... 66
4.1.2. Model Parameters ................................................................................................... 67
4.1.3. Numerical Methods ................................................................................................. 69
4.2. Results ............................................................................................................................ 70
4.2.1. Interstitial Fluid Field ............................................................................................. 70
4.2.2. Single Administration Mode ................................................................................... 71
4.2.2.1. Infusion Duration .................................................................................................. 71
4.2.2.2. Effect of Dose Level ............................................................................................. 76
4.2.3. Multiple Administration Mode ............................................................................... 78
4.2.3.1. Continuous Infusion .............................................................................................. 79
4.2.3.2. Bolus Injection ...................................................................................................... 83
4.2.4. Comparison of 2-D and 3-D Models ...................................................................... 86
4.3. Discussion ...................................................................................................................... 87
4.4. Summary ........................................................................................................................ 89
8
5. Effect of Tumour Properties on Drug Delivery ................................................................. 90
5.1. Heterogeneous Vasculature Distribution ....................................................................... 91
5.1.1. Model Geometry ..................................................................................................... 91
5.1.2. Model Parameters ................................................................................................... 92
5.1.3. Numerical Methods ................................................................................................. 93
5.1.4. Results ..................................................................................................................... 94
5.1.4.1. Interstitial Fluid Field ........................................................................................... 94
5.1.4.2. Doxorubicin Concentration ................................................................................... 95
5.1.4.3. Doxorubicin Cytotoxic Effect ............................................................................... 97
5.1.4.4. Comparison of Various Administration Schedules ............................................... 98
5.2. Tumour Size ................................................................................................................... 99
5.2.1. Model Geometry ..................................................................................................... 99
5.2.2. Model Parameters ................................................................................................. 100
5.2.3. Results ................................................................................................................... 101
5.2.3.1. Interstitial Fluid Field ......................................................................................... 101
5.2.3.2. Doxorubicin Concentration ................................................................................. 103
5.2.3.3. Doxorubicin Cytotoxic Effect ............................................................................. 107
5.3. Discussion .................................................................................................................... 108
5.4. Summary ...................................................................................................................... 110
6. Thermo-Sensitive Liposome-Mediated Drug Delivery to Solid Tumour....................... 111
6.1. Acoustic Model for High Intensity Focused Ultrasound ............................................. 112
6.2. Thermo-Sensitive Liposome-Mediated Delivery ......................................................... 113
6.2.1. Model Description ................................................................................................ 113
6.2.1.1. Model Geometry ................................................................................................. 113
6.2.1.2. Model Parameters ............................................................................................... 114
6.2.1.3. Numerical Methods ............................................................................................. 119
6.2.2. Results and Discussion ......................................................................................... 120
6.2.2.1. Temperature Profile ............................................................................................ 121
6.2.2.2. Drug Concentration ............................................................................................. 123
6.2.2.3. Doxorubicin Cytotoxic Effect ............................................................................. 130
9
6.3. Comparison of 2-D and 3-D Models ............................................................................ 132
6.4. Summary ...................................................................................................................... 135
7. Conclusion ........................................................................................................................... 136
7.1. Main Conclusion .......................................................................................................... 137
7.2. Limitations ................................................................................................................... 138
7.2.1. Tumour Property ................................................................................................... 139
7.2.2. Drug Transport Model .......................................................................................... 139
7.2.3. HIFU Heating Model ............................................................................................ 140
7.3. Suggestions for Future Work ....................................................................................... 140
7.3.1. Realistic Blood Vessel Geometry ......................................................................... 140
7.3.2. Nutrient Transport ................................................................................................. 141
7.3.3. Model Validation .................................................................................................. 141
7.3.4. Type / Patient-Specific Tumour ............................................................................ 141
7.3.5. Chemotherapy Drugs ............................................................................................ 142
7.3.6. HIFU Heating Strategy ......................................................................................... 142
Appendix: Publications during the Project ........................................................................... 143
Appendix: Nomenclature ........................................................................................................ 144
Bibliography .............................................................................................................................. 149
10
List of Figures
Figure 1-1. World cancer death 2012 ........................................................................................... 18
Figure 2-1. Normal and tumor vasculature. .................................................................................. 26
Figure 2-2. Typical architecture of blood vessels in solid tumour ............................................... 27
Figure 2-3. Regions in Solid Tumour ........................................................................................... 28
Figure 2-4. Schematic diagram of drug transport ......................................................................... 34
Figure 2-5. Schematic diagram of three-compartment model ...................................................... 35
Figure 3-1. Drug transport with (a) continuous infusion of drug in its free form, and (b) thermo-
sensitive liposome-mediated drug delivery ................................................................ 45
Figure 3-2. Schematic representation of heat transfer under HIFU heating in (a) tumour and (b)
blood........................................................................................................................... 51
Figure 3-3. Dimensions and coordinates of a spherical radiator .................................................. 53
Figure 3-4. Estimation of doxorubicin diffusion coefficient in tumour tissue.. ........................... 56
Figure 4-1. Model geometry: (a) MR image of the prostate tumour (in red) and its surrounding
tissue (in pule blue); (b) the reconstructed 2-D geometry. ........................................ 67
Figure 4-2. Numerical Procedures. ............................................................................................... 69
Figure 4- 3. Interstitial fluid pressure in tumour and surrounding normal tissue. ........................ 70
Figure 4-4. Doxorubicin concentration in plasma as a function of time after start of treatment, for
bolus injection and continuous infusion of various indicated durations (dose = 50
mg/m2). ....................................................................................................................... 72
Figure 4-5. Spatial mean (a) free and (b) bound doxorubicin extracellular concentration in
tumour as a function of time under bolus injection and different infusion durations
(dose = 50 mg/m2). ..................................................................................................... 73
11
Figure 4-6. Spatial mean (a) free and (b) bound doxorubicin extracellular concentration in
normal tissue as a function of time under bolus injection and different infusion
durations. (dose = 50 mg/m2). .................................................................................... 73
Figure 4-7. Spatial distribution of extracellular concentration of free doxorubicin in tumour and
normal tissues (2-hour infusion, time = 0.5 hour). .................................................... 74
Figure 4-8. Doxorubicin intracellular concentration as a function of time, for bolus injection and
continuous infusions of various durations (dose = 50 mg/m2). ................................. 74
Figure 4-9. Predicated percentage tumour cell survival under bolus injection and continuous
infusions of various durations (dose = 50 mg/m2). .................................................... 75
Figure 4-10. Spatial mean doxorubicin concentration in tumour as a function of time under 2-
hour infusion duration.(a) doxorubicin blood concentration, (b) free doxorubicin
extracellular concentration, (c) bound doxorubicin extracellular concentration, and (d)
intracellular concentration. ......................................................................................... 77
Figure 4-11. Predicated percentage of tumour cell survival under various doses. ....................... 78
Figure 4-12. Spatial mean doxorubicin concentration in normal tissue as a function of time under
2-hour infusion duration (for a 70 kg patient). (a) free doxorubicin extracellular
concentration, (b) bound doxorubicin extracellular concentration. ........................... 78
Figure 4-13. Doxorubicin intravascular concentration under single and multiple continuous
infusions as a function of time (total dose = 50 mg/m2). ........................................... 79
Figure 4-14. Spatial mean doxorubicin extracellular concentration as a function of time under
single and multiple continuous infusion. (a) free doxorubicin, and (b) bound
doxorubicin (total dose = 50 mg/m2). ........................................................................ 80
Figure 4-15. Time course of doxorubicin intracellular concentration under single and multiple
continuous infusions. (total dose = 50 mg/m2) .......................................................... 80
Figure 4-16. Time course of tumour survival fraction under single and multiple continuous
infusions. .................................................................................................................... 81
Figure 4-17. Spatial mean doxorubicin extracellular concentration in normal tissue as a function
of time under single and multiple continuous infusion against time. (a) free
doxorubicin, (b) bound doxorubicin (total dose = 50 mg/m2). .................................. 82
Figure 4-18. Comparison of various doses under single and multiple-infusions. (a) intracellular
concentration and (b) cell survival fraction under single infusion with 50 mg/m2 and
12
multiple infusions with 25 mg/m2 as a function of time ; (c) intracellular
concentration and (d) cell survival fraction under single infusion with 75 mg/m2 and
multiple infusions with 50 mg/m2 as a function of time. ........................................... 83
Figure 4-19. Comparison of doxorubicin concentration under single and multiple bolus injections
as a function of time (total dose = 50 mg/m2). (a) intravascular concentration, (b) free
doxorubicin extracellular concentration, (c) bound doxorubicin extracellular
concentration, and (d) intracellular concentration ..................................................... 84
Figure 4-20. Tumour survival fraction as a function of time with single and multiple bolus
injection ...................................................................................................................... 85
Figure 4-21. Comparison of doxorubicin concentration under single and multiple bolus injections
as a function of time (total dose = 50 mg/m2). (a) free and (b) bound doxorubicin
concentration in normal tissue. .................................................................................. 85
Figure 4-22. Comparison of survival fraction between single continuous infusion and 8 times
bolus injection. ........................................................................................................... 86
Figure 4-23. 3-D model geometry. ............................................................................................... 86
Figure 4-24. Comparison of drug concentration and cell density between 2-D and 3-D models. (a)
free doxorubicin extracellular concentration in tumour as a function of time, (b)
maximum percentage differences in drug concentration and tumour cell density
(total dose = 50 mg/m2) .............................................................................................. 87
Figure 5-1. Model geometry: (a) MR image of the liver tumour (in red) and its surrounding tissue
(in pule blue); (b) the reconstructed 2-D geometry. ................................................... 92
Figure 5-2. Interstitial fluid pressure distribution in tumour and normal tissues.......................... 95
Figure 5-3. Spatial distribution of free doxorubicin extracellular concentration in tumour regions
(total dose = 50 mg/m2, infusion duration = 2 hours) ................................................ 95
Figure 5-4. Spatial mean doxorubicin extracellular concentration in tumour as a function of time.
(a) Free doxorubicin, and (b) bound doxorubicin. (total dose = 50 mg/m2, infusion
duration = 2 hours) ..................................................................................................... 96
Figure 5-5. Spatial mean of doxorubicin intracellular concentration in tumour regions as a
function of time. (total dose = 50 mg/m2, infusion duration = 2 hours) .................... 96
13
Figure 5-6. Spatial mean percentage of tumour cells in tumour regions as a function of time (total
dose = 50 mg/m2, infusion duration = 2 hours). ........................................................ 97
Figure 5-7. Spatial distribution of survival cell fraction in tumour regions (total dose = 50 mg/m2,
infusion duration = 2 hours) ....................................................................................... 98
Figure 5-8. Maximum difference among tumour regions over treatment with various doses. (a)
free doxorubicin extracellular concentration and (b) intracellular concentration
(infusion duration = 2 hours) ..................................................................................... 98
Figure 5-9. Maximum difference among tumour regions over treatment with various infusion
durations. (a) free doxorubicin extracellular concentration and (b) intracellular
concentration (total dose = 50 mg/m2) ....................................................................... 99
Figure 5-10. Model geometry: (a) case_1; (b) case_2; (c) case_3 .............................................. 100
Figure 5-11. Estimation of blood vessel surface area to tissue volume ratio of tumours with
various tumour sizes................................................................................................. 100
Figure 5-12. Interstitial fluid pressure distribution in tumour and normal tissue ....................... 102
Figure 5-13. The spatial mean interstitial fluid pressure and transvacular flux per tumour volume
as a function of tumour size ..................................................................................... 102
Figure 5-14. Free and bound doxorubicin concentration in blood after administration as a
function of time. (infusion duration = 2 hours, total dose = 50mg/m2) ................... 103
Figure 5-15. Spatial mean doxorubicin extracellular concentration as a function of time under 2-
hour continuous infusion, dose = 50mg/m2. (a) Free and (b) bound doxorubicin in
tumour, (c) bound and (d) bound doxorubicin in normal tissue .............................. 104
Figure 5-16. Doxorubicin exchange between blood vasculature and interstitial fluid by
convection as a function of time. (a) free and (b) doxorubicin exchange (2-hour
infusion, total dose =50 mg/m2) ............................................................................... 105
Figure 5-17. Doxorubicin exchange between blood vasculature and interstitial fluid as a function
of time in each tumour. (a) free and (b) bound doxorubicin exchange by diffusion
owing to concentration gradient (2-hour infusion, total dose = 50 mg/m2) ............. 105
Figure 5-18. Comparison of (a) free and (b) bound doxorubicin exchange between blood and
interstitial fluid by convection and diffusion in case_2 (2-hour infusion, total dose =
50mg/m2) .................................................................................................................. 107
14
Figure 5-19. Temporal profiles of predicated intracellular concentration in tumours with different
sizes .......................................................................................................................... 107
Figure 5-20. Survival cell fraction of tumour as a function of time in three tumours with different
sizes (2-hour infusion, total dose = 50 mg/m2) ........................................................ 108
Figure 6-1. Spatial distribution of acoustic pressure .................................................................. 112
Figure 6-2. Measured and computed acoustic pressure profiles at 1.33 MHz and 1.45 MPa
pressure at the focal point. (a) along axis and (b) radial distance.. .......................... 112
Figure 6-3. Model geometry and locations of focus regions. ..................................................... 113
Figure 6-4. Permeability of free doxorubicin as a function of temperature... ............................. 116
Figure 6-5. Fold increase in permeability of liposome-mediated doxorubicin as a function of
temperature............................................................................................................... 116
Figure 6-6. Viscosity of water as a function of temperature....................................................... 117
Figure 6-7. Drug release at (a) 37oC and (b) 42
oC. ..................................................................... 119
Figure 6-8. Numerical Procedure ................................................................................................ 120
Figure 6-9. Variations of temperature in tumour, blood and normal tissue as a function of time
for 10 minutes heating. (a) Maximum and (b) spatial-mean temperature. .............. 121
Figure 6-10. Spatial distribution of temperature in (a) tissue and (b) blood at 10min after heating
starts. ........................................................................................................................ 122
Figure 6-11. Variations of temperature in tumour, blood and normal tissue as a function of time
for various heating durations. (a) Maximum and (b) spatial-mean temperature for 30
minutes heating; (c) Maximum and (d) spatial-mean temperature for 60 minutes
heating. ..................................................................................................................... 123
Figure 6-12. Spatial-mean concentration of encapsulated doxorubicin in blood stream as a
function of time (dose = 50 mg/m2). (a) in tumour, and (b) in normal tissue. ......... 124
Figure 6-13. Encapsulated doxorubicin concentration in blood at 10 min after heating ............ 124
Figure 6-14. Spatial-mean concentration (a) and flux (b) of encapsulated doxorubicin in tumour
extracellular space as a function of time. ................................................................. 125
Figure 6-15. Spatial-mean concentration of encapsulated doxorubicin in extracellular space of
normal tissue as a function of time. ......................................................................... 126
15
Figure 6-16. Spatial distribution of encapsulated doxorubicin extracellular concentration in
tumour and surrounding normal tissues at various time points with 60 minutes
heating. ..................................................................................................................... 126
Figure 6-17. Doxorubicin extracellular concentration with various heating duration as a function
of time. (a) free and (b) doxorubicin concentration in tumour, and (c) free and (d)
doxorubicin concentration in normal tissue. ............................................................ 127
Figure 6-18. Doxorubicin extracellular concentration with various heating duration as a function
of time. (a) free and (b) doxorubicin concentration in tumour, and (c) free and (d)
doxorubicin concentration in normal tissue ............................................................. 128
Figure 6-19. Spatial distribution of doxorubicin concentration and cell survival fraction in
tumour at 10 minutes................................................................................................ 129
Figure 6-20. Doxorubicin intracellular concentration with heating and no-heating as a function of
time........................................................................................................................... 130
Figure 6-21. Survival cell fraction in treatment with various heating duration as a function of
time........................................................................................................................... 131
Figure 6-22. Improvement by HIFU heating. (a) intracellular concentration, and (b) cell killing
.................................................................................................................................. 131
Figure 6-23. Comparison of maximum temperature of tumour in 2-D and 3-D model. ............ 133
Figure 6-24. Comparison of spatial mean temperature in 2-D and 3-D model. (a) tumour and (b)
normal temperature. ................................................................................................. 133
Figure 6-25. Comparison of spatial mean temperature in 2-D and 3-D model. (a) blood
temperature in tumour, and (b) blood temperature in normal tissue. ....................... 134
16
List of Tables
Table 1-1. Processes affected by chemotherapeutic drugs ........................................................... 20
Table 1-2. Common side effects caused by chemotherapeutic drugs ........................................... 20
Table 2-1. Morphological features of tumour microvasculature .................................................. 27
Table 2-2. Mean blood flow rate in human tumour and normal tissue ......................................... 29
Table 3-1. Parameters for tumour and normal tissues. ................................................................. 62
Table 3-2. Parameters for liposome .............................................................................................. 63
Table 3-3. Parameters for doxorubicin. ........................................................................................ 64
Table 4-1. MR imaging parameters .............................................................................................. 66
Table 5-1. MR imaging parameters .............................................................................................. 91
Table 5-2. The relative and scaled values of surface area of blood vessel per tumour size ......... 93
Table 5-3. MR imaging parameters .............................................................................................. 99
Table 5-4. Tumour size and the blood vessel surface area to tissue volume ratio in each model
..................................................................................................................................................... 101
Table 6-1. Acoustic and thermal properties ................................................................................ 114
Table 6-2. HIFU transducer parameters...................................................................................... 114
Table 6-3. Release rates at various temperatures ........................................................................ 119
1. Introduction
Cancer is a major cause of mortality and morbidity in the developed and developing
countries. Currently, cancer is treated primarily by surgery, radiation therapy and
chemotherapy. Novel approaches, such as gene therapy, immunotherapy and
antiangiogenesis therapy, have been developed extensively and are likely to become
clinically available in the next few decades. Except for surgery, effective delivery of
therapeutic agents to tumour cells is the key to success for these therapies. Therefore,
in this thesis, the essential physical and biochemical processes involved in drug
delivery to solid tumours are studied. This chapter presents the relevant background
and preliminaries of the project. The objectives and strategies of the research are then
defined, which is followed by an outline of the thesis.
1. INTRODUCTION
18
1.1. Background
1.1.1. Cancer
Cancer is a set of diseases characterized by unregulated cell growth leading to invasion of
surrounding tissues and spread to other parts of the body [1]. Cancer cells can divide and grow
uncontrollably, forming malignant tumours and finally invading nearby normal tissues. Up to
now, over 200 types of cancer have been discovered. As shown in Figure 1-1, the high mortality
rate associated with cancer makes it one of the most serious life-threatening diseases in the world.
Around 159,000 people died from cancer in the UK in 2011, indicating 252 deaths for every
100,000 people.
116.0+
99.6 – 116.0
89.7 – 99.6
73.3 – 89.7
<73.3
No Data
Death per 100,000
Population*
*All cancers except non-melanoma skin cancer
Figure 1-1. World cancer death 2012
The causes of cancer are diverse, complex and not totally understood [2]. Various factors have
been suggested to increase the risk of cancer, including pollution, radiation, industrial products
and certain types of infections caused by damaged genes or combined with existing genetic faults
within cells. Cancerous mutations of proliferation related genes, proto-oncogenes and/or tumour
suppressor genes, drive cancer cells to go on duplicating themselves.
Cancer seems to develop progressively [2]. As cancer cells expand, the original space becomes
insufficient for their uncontrolled growth, resulting in invasion of increasing number of cells to
the adjacent normal tissues. In order to acquire oxygen and nutrition to support cancer growth,
normal cells are signalled to help develop new blood vessels. These vessels tend to be leaky and
with abnormal geometric properties, which are different from blood vessels in normal tissues. In
1. INTRODUCTION
19
addition, cancer cells can migrate from the primary cancer and spread to more distant parts of the
body through the blood stream or lymphatic system to establish new colonies of cancer cells.
These new settlements, termed metastases, are the major cause of death since vital organs can be
damaged by cancer cells [1]. Ideally, treatments should start before cancer cells invade into other
organs.
1.1.2. Anticancer Therapy
Various anticancer treatments are available in clinics, and the choice of these treatments is
strongly dependent on the location, growth stage of cancer as well as the health condition of
patients, etc. The main anticancer therapies are outlined below.
(1) Surgery
Surgery is cutting away tissue from patient’s body. Given that it is a local treatment, surgery may
cure a localised cancer that is totally contained in one area before metastases taking place. In
surgery, tumour and some of its surrounding normal tissues are removed to make the margin
clear. Considering that cancer may invade the lymphatic system, the nearest lymph nodes are
also removed. This is a preferred treatment for cancer, and can offer a complete cure in some
cases [1].
After spreading of cancer cells, surgery usually cannot cure the cancer, but may also help
patients live longer and reduce/control pain. Another application of surgery is to prevent or
reduce the risk of cancer. People at a high risk of a particular type of cancer (e.g. bowel or breast
cancer) may choose surgery to remove the relevant tissue.
Surgery is limited by the type of cancer. Certain types of cancer are difficult to be removed due
to their location, and insufficient removal may result in regrowth of cancer. On the other hand,
spalling of cancer cells during surgery may cause the spread of cancer cells.
(2) Chemotherapy
Chemotherapy is the use of cytotoxic or cytostatic drugs that respectively kill malignant cells or
prevent them from proliferating. The effectiveness of chemotherapy strongly depends on the
1. INTRODUCTION
20
amount of drugs accumulated in the tumour, exposure duration to drugs and the proportion of
cell population that is proliferating.
There are more than 100 different anticancer drugs currently available and new drugs are being
developed all the time. The choice of drugs is highly related to the type and stage of tumour.
Chemotherapy kills tumour cells by damaging the genes that make cells divide, or it may
interrupt the chemical process involved in cell division. The process affected by normal
chemotherapy drugs are summarized in Table 1-1.
Table 1-1. Processes affected by chemotherapeutic drugs [1]
Drug Process Affected Cancer Treated
Cisplatin DNA synthesis Ovary, Testis
Methotrexate Dihydrofolate reductase Breast, Placenta
5-Fluorouracil Thymidylate synthase, RNA synthesis Stomach, Breast, Colorectal
Doxorubicin DNA & RNA synthesis Lung, Breast, Leukaemia
Vincristine Tubulin polymerisation Lymphoma, Testis
Chemotherapy is a systemic treatment. Drugs are circulated around the body in the bloodstream
and can reach cells almost anywhere. Without specifying the target, normal cells can also be
killed and hence the normal function of tissue can be compromised. This is known as side effects,
which can seriously offset the treatment outcome. Side effects vary according to the drug, but
there are some most common side effects, as shown in Table 1-2.
Table 1-2. Common side effects caused by chemotherapeutic drugs [1]
Drug Side Effects
Cisplatin Nephrotoxicity, Ototoxicity, etc
Methotrexate Myelosuppression, Alopecia, Dermatitis, etc
5-Fluorouracil Myelosuppression, Cardiac toxicity, etc
Doxorubicin Cardiomyopathy, Myelosuppression, Leukopenia, etc
Vincristine Chemotherapy-induced peripheral neuropathy, Hyponatremia, etc
During chemotherapy, drugs often stop the bone marrow from generating enough blood cells.
Without adequate white blood cells, risks of developing infections will increase, which low
levels of red blood cells result in the shortage of nutrition and oxygen supply, causing
1. INTRODUCTION
21
breathlessness and tiredness. Most drugs cause hair loss, but this effect is temporary and hair
starts to grow back a few weeks after chemotherapy ends.
Changes in the way that liver, kidney, lung and heart work can also be caused by chemotherapy
drugs. These changes are usually temporary, but can be permanent for some patients.
Doxorubicin is a typical anticancer drug, and its most serious side effect is cardiomyopathy,
which can lead to heart failure. In order to minimise the risk for this, the total dose of
doxorubicin in a life cycle is limited to 550 mg per unit body surface area [3].
(3) Radiotherapy
Radiation is applied to kill malignant tumour cells by causing irreversible damages to DNA
within cancer cells. Since tumour stem cells are not as sensitive as cycling cells to radiation, the
entire treatment process may last longer than other types of treatment. According to the way
radiation is used, basically there are two types of radiotherapy: external radiotherapy and internal
radiotherapy.
In external radiotherapy, a narrow radiation beam with high intensity is generated to a well
identified area of the patient’s body. X-ray and gamma ray are the most common radiation used
clinically. Since radiation can pass through the body, superficial and deep-seated tumours can be
treated without surgical trauma. Given that the treated area is well identified before radiotherapy,
damage to adjacent and overlying normal cells can be minimized by re-directing the radiation
beam to restrict the radiation dose only being focused in tumour.
In internal radiotherapy, radioactive implants are inserted within or close to the target tumour,
and a radioactive liquid is administrated orally or injected first, which is then transported by the
circulation system. The radioactive component of the liquid is called isotope, which is designed
to attach to a 'bioconjugate' molecule, such as an antibody, to cause the liquid to be taken up by
tumour cells.
Although radiotherapy is a local treatment, some normal cells are inevitably killed, resulting in
undesirable side effects. During the treatment, tiredness is a common side effect since healthy
cells need to be repaired or the level of blood cells drops. In addition, radiotherapy can also
cause side effects in particular organs, such as in the head, neck or chest.
1. INTRODUCTION
22
(4) Novel and Emerging Therapies
High intensity focused ultrasound (HIFU) is a novel and fast developing therapy [4]. High
frequency ultrasound waves are emitted from an ultrasound transducer, and the waves deliver a
strong beam to the targeted region in order to elevate the temperature to kill tumour cells. It is a
highly targeted and localised treatment. Because of the non-invasive nature of the sound wave
used, the associated side effect is less than the other treatment methods. Patients may feel
localised pain in the first 3~4 days.
Photodynamic therapy (PDT) is usually used to treat non-melanoma skin cancer [5]. Drugs
(known as photosensitiser) are firstly administrated to add a specific type of light sensitivity to
tumour cells, which are then killed by exposure to this light. However, PDT can make skin and
eyes sensitive to light, and may cause death of nearby normal cells.
Gene therapy involves inserting specific genes into malignant or normal cells in order to modify
their gene expressions. Either viral or non-viral agents can be used to deliver the specific genes,
and viral must be attenuated before used. This therapy is with high specificity, but is still in
development stage. A number of early stage clinical trials have been carried out [6-8].
Clinically, the treatment methods are strongly dependent on the type, location and growth stage
of tumour as well as the health condition of patient. Usually, various treatments are combined in
order to improve the overall treatment outcome. Most of these therapies depend on effective
delivery of therapeutic agents to the targeted tumour.
1.1.3. Thermo-Sensitive Liposome-Mediated Drug Delivery
For an effective delivery, adequate therapeutic agents should reach the tumour while drug
accumulation in normal tissues needs to be prevented or reduced in order to minimize side
effects. Since most therapeutic agents are carried by systemic bloodstream to the entire body,
their usages are hence limited. As an improvement, thermo-sensitive liposome-mediated drug
delivery has been developed to overcome this shortcoming.
Anticancer drugs are encapsulated by thermo-sensitive liposomal particles which are
administrated into the blood stream as usual [9]. Ideally, these particles should retain their load at
body temperature, and rapidly release the encapsulated drugs within a locally heated tumour
1. INTRODUCTION
23
region. The fast release is due to liposome thermo-responsiveness, which arises from a phase
transition of the constituent lipids and the associated conformational variations in the lipid
bilayers [10-12].
Localised heating can be achieved by various means, including: microwave, radiofrequency
electric current, laser, and HIFU transducer [13]. The sound wave generated by HIFU transducer
is a mature technology which has been adopted in HIFU treatment alone without the use of drugs,
and is found to have very few side effects.
1.2. Objectives of the Study
The delivery of anticancer drugs consists of multiple processes, as the drugs must make their
way into the blood vessels of the tumour, pass through the vessel wall to enter the interstitial
space and be taken up by tumour cells. On the one hand, these processes are dependent on the
physicochemical properties of the drugs, such as diffusivity and drug binding to cellular
macromolecules. On the other hand, the biological properties of a tumour, including tumour
vasculature, extracellular matrix components, interstitial fluid pressure, tumour cell density and
tissue structure and composition, also serve as determinants in these processes. Unfortunately,
the structure and morphology of microvessels in malignant tumours is highly abnormal and
heterogeneous, which create a great barrier for effective delivery of drugs to cancer cells.
The overall aim of this project is to develop a multi-physics model for the simulation of transport
processes of therapeutic agents in solid tumours. The main focus of this project is on the
development of transport models for free drug and drug loaded thermal-sensitive liposomes, with
the following specific objectives:
(1) To develop a mathematical model for the transport of anticancer drugs in their free form;
(2) To determine the optimal administration mode and schedule for maximum accumulation of
drugs in tumour cells;
(3) To develop a mathematical model for the transport of drug-loaded thermosensitive liposomes;
(4) To fulfil objective (3), a bioheat transfer model with HIFU heating needs to be developed;
(5) To examine the effect of different drug delivery modes on drug distribution in solid tumours.
1. INTRODUCTION
24
A research strategy has been developed in order to accomplish these objectives. The
mathematical models are applied to realistic tumour models reconstructed from magnetic
resonance images (MRI). In the basic tumour models, a uniform microvasculature is assumed.
This assumption is subsequently relaxed by incorporating microvasculature heterogeneity
derived from in vivo MRI data, and the effect of microvasculature distribution on drug
concentration in tumour is investigated. Finally, the drug delivery model is coupled with the
bioheat transfer model under HIFU heating in order to study thermal-sensitive liposome-
mediated drug delivery.
1.3. Structure of This Thesis
This thesis includes seven chapters, which are organized in a logical sequence. In Chapter 2, a
comprehensive survey of literature is given for the research to cover tumour properties, drug
delivery modes and drug transport processes. In Chapter 3, mathematical models utilized to
achieve the objectives and their implementations are described. Chapters 4, 5 and 6 present the
findings and analysis of results of drug transport process and treatment efficacy with various
administration modes and biological properties. Chapter 4 compares the bolus injection and
continuous infusion, and Chapter 5 investigates the influence of heterogeneous vasculature and
the size of tumour. Chapter 6 focuses on drug transport under thermal-sensitive liposome-
mediated drug delivery. Finally, conclusions and suggestions for future work are outlined in
Chapter 7.
2. Literature Review
In this chapter, a comprehensive review of the literature on topics related to drug
transport in solid tumour is given. Firstly, the unique characteristics of solid tumour
are described, which serves as an introduction to the microenvironment in tumour.
This is followed by a review of previous studies of drug delivery modes, covering
both medical and mathematical modelling aspects. Finally, the key processes
involved in drug delivery to solid tumour are summarized.
2. LITERATURE REVIEW
26
2.1. Characteristics of Solid Tumour
Different from normal tissues, solid tumours have abnormal characteristics that may strongly
influence the transport of anticancer drugs to and within tumours.
2.1.1. Tumour Vasculature
Microcirculation plays an important role in the delivery of anticancer agents to solid tumours
[14-16]. Small tumours with a diameter less than 2 mm are perfused by blood from their host
normal tissues [17]. With the growth of a tumour, new blood vessels are formed either through
recruitment from pre-existing network of the host tissue or resulting from the angiogenic
response of tumours [18, 19]. Both morphological features and functions of tumour vasculature
differ from those in normal tissues.
Figure 2-1. Normal and tumor vasculature.
A) Normal vasculature: vessels are aligned in a well-organized manner; B) Tumor vasculature: vessels are dilated
with non-uniform diameters and random branching patterns (extracted from [20])
Tumour vasculature becomes tortuous, elongated, and often dilated as a result of tumour growth
(Figure 2-1). The morphological structure of vessels can also change in the growing process of a
tumour, with some features that are rarely present in normal tissues. As shown in Figure 2-2,
trifurcation is a typical pattern which is frequently seen in tumour vasculature. Loops can be
found in both arterial and venous trees in tumours, and there are two types: self-loop and true
loop [21]. Self-loops are planar loops consisting of only two vessels without any other side
branches, whereas true loops are non-planar composed of several vessel segments with many
branches. Polygonal structure exists in the tumour capillary meshwork. Venous convolution is
2. LITERATURE REVIEW
27
often observed in close proximity to the feeding/draining vessels. The result of this kind of
vascular architecture may lead to plasma skimming and heterogeneous vessel haematocrit levels.
Figure 2-2. Typical architecture of blood vessels in solid tumour (extracted from [21] )
Including (1) trifurcations, (2) self-loop, (3) characteristic polygonal structure of the capillary meshwork, (4) venous
convolutions, and (5) small (20~40 µm) vessels branching off of large (200 µm) vessels
The diameters of arterioles and venules decrease monotonically as branching order increases.
Comparisons presented in Table 2-1 [21-24] show that blood vessels in tumour have a large
diameter, which may exceed the size of anticancer agents (<1~2 µm for free drugs and 100 nm
for liposome particles [25, 26]). Comparisons between unit volume of tumour and normal tissue
show that the length, surface area and fraction of vasculature volume are significantly larger in
tumours.
Table 2-1. Morphological features of tumour microvasculature
Diameter (µm) Length Surface Area Vascular Volume
Fraction (%) Capillary Venules (cm/mm3) (mm
2/mm
3)
In Tumour 5~20 15~70 36 70 50
In Normal Tissue 5~8 12~50 160 20 20
One of the unique characteristics of tumour vasculature is its high leakiness. The abnormal
structure of vessel wall in tumours [27-29] may be one of the reasons. Large inter-endothelial
junctions, increased numbers of fenestrations, vesicles and vesico-vacuolar channels, and a lack
of normal basement membrane are often found in tumour vessels [30, 31]. Perivascular cells
have abnormal morphology and heterogeneous association with tumour vessels. In
correspondence to these structural alterations in the tumour vessel wall, the cut-off size of pores
on tumour microvasculature is in the range of 100~780 nm, varying according to the location and
growth stage of the tumour [26, 32, 33]. Compared to the pore size of 2~6 nm in most normal
2. LITERATURE REVIEW
28
tissues [34-36] and 40~150 nm in kidney, liver and spleen [37, 38], tumour vasculature with
large pores is favourable for anticancer drugs to pass through, and thereby easing the drug
delivery process.
The distribution of tumour vasculature depends strongly on the location and growth stage of the
tumour. In theory, four regions can be categorized based on the type and function of tumour
vasculature [39]: (1) necrotic region with no functional vasculature; (2) semi-necrotic regions
with capillary, pre- and post-capillary extended; (3) stabilized microcirculation region
characterised mainly by venules, veins and a few arterioles; and (4) tumour advance front with
arterioles, capillaries and venules. Jain simplified this region-categorization as necrotic region,
semi-necrotic region and well-vascularised region along the radial distance [40], as shown in
Figure 2-3.
Necrotic
Region
Semi-
Necrotic
Region
Well
Vascularized
Region
Figure 2-3. Regions in Solid Tumour (extracted from [40] )
In reality, two types of vascular structure can be observed from in vivo images [41]. In tumours
with peripheral vascularisation, the centre is poorly perfused and vessels are located mainly in
the periphery. These may also occur in fully grown tumours that have already invaded into
normal tissues. In tumours with central vascularisation, vessels proliferate from the centre and
form a tree-like structure. In this case, vessel volume in the centre is higher than that in the
periphery of the tumour.
2.1.2. Blood Flow Rate
Given that anticancer agents are carried by the blood stream to the target solid tumour, tumour
blood flow rate may greatly influence the transport of drugs. Under the assumption of steady and
2. LITERATURE REVIEW
29
fully developed flow and ignoring geometric complexities, the volumetric flow rate in a tumour
vascular network is proportional to the pressure difference between its arterial and venous ends,
and inversely proportional to blood viscosity and geometric resistance [42].
Microvasculature pressure at the arterial ends is found to be similar in both tumour and normal
tissue [43], but pressure in the venular side is significantly lower in a tumour when compared to
normal tissue [17]. Because most blood vessels in tumours are veins whilst arteries mainly exist
in tumour advance front region [41], the driving force for blood flow is greater at the periphery.
This is one possible reason for the heterogeneous distribution of blood flow in tumour [39, 44].
Viscous resistance depends on plasma viscosity and parameters describing the number, size and
rigidity of red blood cells, while geometric resistance is directly associated with the
morphological features of microvessels, such as the number of vessels, branching patterns,
diameter and length [42]. Both of these resistances may be altered by changes in environmental
or internal conditions. Compared to normal tissue, greater resistance is found in tumours owing
to the drainage force from the tumour cells and proteins existing in the extracellular space, as
well as the large vessel diameter and long vessel length.
Table 2-2. Mean blood flow rate in human tumour and normal tissue (ml/g/min) [45].
Brain Lymphatic Breast Uterus Liver
Tumour
0.50 0.45 0.23 0.33 0.12
Normal Tissue 0.57 0.40 0.83 0.13 1.05
A number of studies of blood flow in human tumours and their holding tissues have concluded
that there is a wide range of variations in blood flow rate even for the same type of tumour [45].
Comparisons summarized in Table 2-2 show that the difference in mean blood flow rate between
tumours and normal tissues depend strongly on their location.
2.1.3. Lymphatic Drainage
The lymphatic system is a part of the circulatory system spreading into many organs. Lymphatic
vessels are structurally similar to capillaries, with a wall formed by a single layer of endothelial
cells [46]. However, the lymphatic wall is more permeable than the capillary wall and is not size-
selective [17]. Its main function is to return interstitial fluid back to the blood stream partly due
2. LITERATURE REVIEW
30
to the low pressure in lymphatics, and to clear extracellular matrix deposits, such as proteins,
from the interstitial space in most normal tissues [47]. The diameter of lymphatic vessels in
tissue is around 40 μm, and the velocity is in the range of 1.4~20 μm/s, with an average value of
8 μm/s [46]. Hence, drugs can be constantly cleared in regions with a high density of functional
lymphatic vessels.
2.1.4. Interstitial Fluid Pressure
The interstitial space of a tumour is complex partly due to the extracellular matrix and non-
uniform vasculature distribution [46]. The pressure in a tumour can exceed atmospheric by
greater than 4200 Pa [46, 48]. This high pressure can be attributed to: (1) insufficient normal
removal of interstitial fluid due to lack of lymphatics; (2) high permeability of tumour
vasculature, and (3) vascular collapse caused by cell proliferation in a confined volume [15].
Models of microenvironment as complex as interstitial space of a tissue require a proper
description of its microstructure. The organization of capillaries in tumours is random owing to
the abnormal morphological features, and the inter-capillary distance is usually 2~3 orders of
magnitude smaller than the length scale for drug transport [49], hence a useful model can be
based on percolation description of porous media governed by Darcy’s law [46].
A theoretical platform has been set by Baxter and Jain [49] for transvascular exchange and
extravascular transport of fluid and macromolecules in tumour based on Darcy’s law describing
the interstitial space as a porous media. Validated by experimental data, their simulations of
tumour microenvironments in both tissue-isolated and subcutaneous tumours indicate that
interstitial fluid pressure in a tumour is higher than in normal tissue with a uniform distribution
in the centre even if in the presence of a necrotic core, and there is a sharp drop of interstitial
pressure at the periphery of the tumour [49-51].
2.2. Modes of Drug Delivery to Solid Tumour
The route and method of drug administrated affect the kinetics of bio-distribution and
elimination, therefore, the effectiveness of chemotherapy [46]. Modern pharmaceutical science
provides many options for drug administration to meet specific requirements in clinical
anticancer treatments.
2. LITERATURE REVIEW
31
2.2.1. Traditional Administration
Anticancer drug administration traditionally includes oral delivery and intravenous injection.
Because eating is one of the most common acts in daily life, oral intake is always preferred. It is
painless, uncomplicated and self-administrated [46]. However, due to the fact that many drugs
are degraded within the gastrointestinal tract, or cannot be sufficiently absorbed, the treatment
efficacy is seriously compromised.
Drug delivered by intravenous infusion is therefore adopted to offer a more rapid and efficient
therapeutic outcome. With this administration method, nearly 100% of the drug is bioavailable
and the plasma concentration level can also be controlled continuously [46]. The disadvantages,
such as risk of overdose and infection, can also be avoided. Nowadays, intravenous injection is
the most common administration mode in anticancer treatment.
Studies on infusion duration were attempted with an aim to maximise intracellular drug
concentration and reduce side effects. Noticing that high plasma concentration of doxorubicin
may cause serious cardiotoxicity [3, 52], Legha [53] prolonged the infusion duration to 48 hours
or 96 hours in treatments for 21 patients, while the other 30 control patients received standard
intravenous injection with the same clinical dose. It was found that plasma levels of doxorubicin
were reduced by continuous infusion, thereby reducing the risk of cardiotoxicity. Comparison
also showed that the anticancer efficacy was not compromised.
Similar clinical studies were carried out by Hortobagyi [54]. The initial group of breast cancer
patients were treated by doxorubicin with bolus injection while in the other group doxorubicin
was administrated via a central venous catheter over a 48-hour or 96-hour continuous infusion
schedule. Anticancer treatment outcomes revealed that there were no differences between these
two groups, but continuous infusion was less cardiotoxic than bolus injection.
El-Kareh and Secomb [55] suggested that shorter infusion duration might improve treatment
efficacy, and explored it by numerical studies. The anticancer efficacy is reflected by survival
cell fraction based on the predicted peak intracellular concentration over the entire treatment
period. Comparison between bolus injection and continuous infusion of 50, 100, 150 and 200
minutes indicated that infusion duration had a strong influence on the predicted outcome. An
optimal duration for continuous infusion was found to be within 1 to 3 hours.
2. LITERATURE REVIEW
32
2.2.2. Nanoparticle-Mediated Delivery
In order to reduce the risk of side effects caused by high drug concentration in normal tissues,
cytotoxic drugs are encapsulated in nanoparticles, typically liposomes, before being
intravenously administrated into blood stream [56, 57]. The term “liposome” was firstly
introduced to describe one or more concentric lipid bilayers enclosing an equal number of
aqueous media which can be entrapped in the core during formation [46, 58]. Usually, liposomes
are designed to be in the range of 70 to 200 nm in diameter in order to increase its circulation
time in the blood stream [59]. Decreasing the diameter of liposome to less than 70 nm would
result in 70% of the injected dose being accumulated in liver, while larger liposomal particles
with a diameter greater than 200 nm are more likely to be taken up by the spleen, resulting in a
reduced level in blood [60].
Because of phagocytosis by cells of the reticuloendothelial system, liposomes disappear rapidly
in the blood stream [46]. Hence, modifications on the liposome structure are required.
Polyethylene Glycol (PEG) is now widely adopted as an important component in liposomal layer
in order to prolong the circulation of liposomal nanoparticles in the blood stream. The drug
concentration, therefore, can be maintained at a sufficient high level for effective cell killing [56,
61]. The pharmacokinetics of liposome-mediated doxorubicin in cancer patients was
qualitatively analysed by comparing with the same dose of doxorubicin delivered in free form
[62]. Results have shown clearly that liposome-mediated delivery can reduce plasma clearance
and increase doxorubicin concentration within tumours. Plasma elimination of doxorubicin-
loaded liposomes follows a bi-exponential decay function with half-lives of 2 and 45 hours [62].
In vivo experiments on human prostate carcinoma, which is implanted subcutaneously into nude
Swiss mice, demonstrated that liposome-encapsulated doxorubicin could enhance the therapeutic
efficacy owing to reduced systemic elimination, increase penetration in tumour and prolong
liposome presence with slow drug release [63]. To assess the clinical benefit of liposome-
encapsulated doxorubicin, 297 patients with metastatic breast cancer with no prior chemotherapy
were randomized to receive 60 mg/m2 liposome-encapsulated doxorubicin or conventional
doxorubicin in combination with 600 mg/m2 of cyclophosphamide until disease progression or
unacceptable toxicity was reached [64]. Results showed that liposome-encapsulated doxorubicin
could significantly reduce cardiotoxicity and improve the therapeutic effectiveness.
2. LITERATURE REVIEW
33
Controlled release is important in making chemotherapy a localised treatment. With a thermo-
sensitive liposomal membrane, in theory, no drug can be released until it is heated beyond its
phase transition temperature [65]. A large number of experimental studies have been carried out
to meet this goal. Thermo-sensitive liposomes sterically stabilized in human plasma have been
developed by Gaber [66], while Unezkazi [67] and Lindner [68] developed improved
formulations to offer both prolonged circulation and thermo-sensitivity. Optimization study on
liposome particle was carried out by Tagami [69] to reduce the release rate at body temperature
while maximising the release rate at mild hyperthermia.
Fast temperature elevation can be obtained by laser, microwaves, radiofrequency electric current
[13] and high intensity focus ultrasound (HIFU) [70]. Ultrasound has been used clinically to
apply thermal therapy non-invasively at targets that are unavailable to other heating methods
[71]. Staruch tested localised drug release with HIFU heating by in vivo experiments on rabbits
[72, 73]. In their experiments, a HIFU beam was scanned in a circular trajectory to heat a given
region in normal thigh [72] and tumour [73] to 43oC lasting for 20~30 minutes. Localised
heating is controlled by MRI thermometry. Encapsulated doxorubicin with thermo-sensitive
liposome was intravenously administrated during hyperthermia. Comparison between heated and
unheated regions indicated that MRI-controlled HIFU hyperthermia could enhance local drug
delivery.
Temperature elevation in a liver tumour was predicated by numerical studies in Sheu’s work [74].
Based on a patient-specific liver model, the large hepatic artery was modelled explicitly to
investigate the impact of blood flow on temperature elevation. However, neither the real
geometry of hepatic tumour nor drug release was included in this study.
Researchers also investigated HIFU heating modes in order to achieve a homogeneous
temperature profile in tumour. Based on a 2-D idealized circular model, variables in fast
scanning method with a short sonication duration for HIFU treatment were examined [75].
Results showed that fast scanning method could produce a planned lesion regardless of its
scanning path. The final temperature profiles were highly dependent on the applied power and
tissue perfusion. Theoretical and experimental studies [76, 77] suggested that signal point fixed-
gain proportional-integral control coupled with multiple heating points along a rapid scanned
trajectory could result in a homogeneous temperature profile in a given region.
2. LITERATURE REVIEW
34
2.2.3. Polymer Implant
Apart from nanoparticle-mediated delivery, a polymer implant is another delivery method
developed to overcome the lack of targeting in common methods. The anticancer drug is firstly
dissolved in polymers, and then the drug needs to diffuse through the polymer after being
implanted and finally dissolve into the extracellular space of tumour. A number of polymeric
materials have now been approved for clinical use. The most extensive type is non-degradable
hydrophobic polymers. The development of biodegradable polymers with the advantage that no
residual material remains in the tissue has enabled more anticancer agents with low diffusivity
to be delivered to solid tumours [46]. However the design of these polymers is complex since
large quantities of potentially harmful products may be released in the body.
Wang [78] showed that implanting IgG-loaded polymer in the cavity left after surgical removal
of a tumour may improve the anticancer efficacy. However, Teo [79] found that surgical removal
of the tumour alone could only enhance drug delivery in bis-chloroethylnitrosourea (BCNU)
treatment in limited time. Intratumoural administration with optimization is capable to improve
treatment efficacy and reduce drug concentration in normal tissue [80, 81].
2.3. Drug Transport in Solid Tumour
Systemic administration is the main delivery method for chemotherapeutic agents. Following its
administration into the blood stream, anticancer drug experiences sequential processes before
reaching the targeted tumour cells. These processes (shown in Figure 2-4) are described in detail
below.
Protein
Cells
Blood
Vessel
Interstitial
Fluid
Drug
Figure 2-4. Schematic diagram of drug transport
2. LITERATURE REVIEW
35
2.3.1. Transport within Blood Vessel
After administration, therapeutic agents begin to circulate in the systemic circulation to all over
the body. Drug transport within blood stream is determined by drug properties and the in vivo
environment. Drugs that are transported via the blood may disperse into tissue when passing by.
In addition, kidney and other organs can eliminate drugs from the circulation system by filtering
plasma continuously, and thereby compromising the treatment effect. This function of renal
excretion is known as plasma clearance, which may vary according to the drug type [82-84].
Rapid Peripheral
Compartment (V2)
Central
Compartment
(V1)
Slow Peripheral
Compartment (V3)
I(t)
Figure 2-5. Schematic diagram of three-compartment model
Pharmacokinetics is classically represented by compartment models describing drug plasma
concentration as a function of time, I(t). A human body can be represented as three
compartments with the volume distribution at steady state shown in Figure 2-5: the central
compartment (V1) for blood plasma, rapid (V2) and slow (V3) peripheral compartments
corresponding to dense and sparse vessel volumes, respectively. Drugs in the central
compartment can be cleared out, and there are two-way exchanges between the central and
peripheral compartments, as shown in Figure 2-5. A two-compartment model has also been
adopted in mathematical modelling of the delivery of anticancer agents, which consists of a
central compartment and periphery compartment, with inter-compartmental exchange taking
place in both directions.
Plasma concentration is associated with plasma clearance, dosage and administration mode. The
pharmacokinetics of doxorubicin has been evaluated in the plasma of 12 breast cancer patients
receiving no prior treatment [85]. Drug concentration was obtained by testing plasma samples
collected at various times after administration. The successive half-lives were 0.08 hour, 0.822
hour and 18.9 hour, indicating that a three-component exponential function would be better
suited for doxorubicin concentration in blood plasma.
2. LITERATURE REVIEW
36
In addition to red blood cells, white blood cells and platelets, blood also contains a range of
proteins. Extensive bindings between these proteins and drugs have been observed in clinical and
experimental conditions. Plasma pharmacokinetics of doxorubicin with a 15 minutes continuous
infusion were studied in 10 breast cancer patients [86]. By applying high performance liquid
chromatography, Green found that 75±2.7% of the total plasma concentration of doxorubicin
was bound with protein over the whole range of observed concentrations. This high rate of
binding can effectively reduce the plasma concentration of free doxorubicin and affect the
amount of drug reaching tumour cells. On the other hand, the retention of drug in blood can be
extended because of the low plasma clearance rate of protein [17].
2.3.2. Transport through Blood Vessels
Being transported within the circulatory system, some drugs can be extravasated to the interstitial
space owing to the high permeability of tumour vasculature. Drug transport through the
vasculature wall depends on convection and diffusion.
Drug transport by convection is associated with fluid flux across the vessel wall. This fluid
movement is determined by hydrostatic and osmotic pressure difference between the lumen and
interstitium. Drug transport by diffusion is driven by the concentration gradient between blood
and interstitial fluid, and is also related to the permeability and area of blood vessels for drug
exchange. This implies that drug can also be transported back to blood when the gradient
reverses. Although drugs can be carried by endothelial cells, transcytosis has been found to play
a minor role when compared with convection and diffusion [17].
The pore model is widely adopted in numerical study of drug transport through blood vessels.
Deen [87] reviewed studies of the hindered transport through pores on membrane that were
aimed at predicting applicable transport coefficients from fundamental information, such as the
size, shape and electrical charge of the solutes and pores in liquid-filled pores. The pioneering
work on drug transport in solid tumour was conducted by Baxter and Jain. Employing the pore
model described in Deen’s work [87], they developed a general mathematical platform for trans-
vascular and interstitial transport of fluid and macromolecules [49, 50, 88, 89].
Through application to an idealized tumour core model, Eikenberry [90] investigated the
delivery of soluble doxorubicin from a capillary to its surrounding tumour mass to examine the
2. LITERATURE REVIEW
37
penetration of doxorubicin under different dosage regimens and microenvironment in solid
tumour. Drug infusion duration was found to be an important factor in determining the spatial
profile of tumour cell killing, and extending the infusion duration or fractionating the dose would
help improve the treatment effect. Numerical results showed that only a limited amount of
doxorubicin could be delivered into solid tumours.
The application of a pore model was extended by El-Kareh and Secomb [55] to study liposome-
mediated delivery of doxorubicin to solid tumours. A mathematical model was developed and
applied to compare the performance of various administration modes of both free form and
liposomal encapsulated doxorubicin. The treatment outcome was predicated based on peak
intracellular concentration without considering the dynamics of tumour cell density. Their
simulation results indicated that continuous infusion with an optimal infusion duration yielded
the best treatment efficacy, and there was a clear advantage of thermo-sensitive liposomal
delivery at some doses but only if the blood was not significantly heated.
2.3.3. Transport in Interstitial Space
Following extravasation, drugs move in the interstitial space to reach tumour cells. The main
pathways for drug movements are also attributed to convection and diffusion, which are highly
dependent on interstitial fluid pressure and interstitial fluid velocity within tumour. Drug
convection in interstitium is determined by hydraulic conductivity and pressure gradient, and
diffusion depends on the concentration gradient and diffusion coefficient. Diffusion within a
tissue is slower than diffusion through a medium filled with interstitial fluid, due to the presence
of cells [46]. The composition of the extracellular space also affects the diffusivity of compounds
through tissue, especially for macromolecules which are mostly influenced by the presence of
proteins in interstitial fluid. The fact that diffusivity also depends on the architecture of tissue [46]
means that the diffusion coefficient for the same drug can be different in tumour and normal
tissue.
In the work of Netti [91], the tumour was modelled as a poroelastic material. The aim was to
determine the mechanisms regulating interstitial fluid pressure and develop methods to overcome
the interstitial fluid pressure barrier in drug delivery. Experimental and numerical results
demonstrated that higher uptake of macromolecular would result from periodic modulation of
2. LITERATURE REVIEW
38
blood pressure rather than acute or chronic increase of intravascular pressure. Hence pressure
elevation caused by drug administration would have a limited effect on drug delivery.
Goh [51] develop a 2-D simulation platform to investigate the spatial and temporal variation of
doxorubicin concentration in hepatoma. The model geometry was reconstructed from patient-
specific CT images, including the necrotic core, viable tumour and its surrounding normal tissue.
Increase in vascular pressure caused by intravenous bolus injection was considered, and the
predicated results showed that interstitial fluid pressure stabilized fairly quickly. The results also
indicated that diffusion was the dominant transport mechanism. It was concluded that
doxorubicin transport to solid tumour was strongly dependent on the vascular exchange area,
doxorubicin cellular metabolism and DNA binding kinetics.
The effect of transient interstitial fluid flow on drug delivery in brain tumour was addressed by
Teo [79]. After surgical removal of the necrotic core, polymer wafers that contained bis-
chloroethylnitrosourea (BCNU) were implanted in the cavity. The mathematical platform
established by Goh [51] was applied to simulate the intra-tumoural release of BCNU and its
penetration through resection cavity to the tumour. The interstitial fluid flow was found to reach
equilibrium in less than 3 hours, and surgery could only enhance drug delivery temporally.
Owing to the presence of a variety of proteins and nucleic acids in the interstitial space,
extensive binding may take place. Experimental studies have shown that binding can delay drug
transport [92-94]. In an experiment on thyroid cancer cells, 5-fluorouracil, cisplatin and inulin
were found to be evenly distributed in cell spheroids within 15 minutes [17, 93, 94], while
doxorubicin, methotrexate and paclitaxel, which bind to macromolecules, remained in the
periphery of spheroids [95-97]. Since a tumour is always surrounded by normal tissues and
interstitial fluid velocity is higher at the periphery of tumour than in the centre, drug exchange
can also take place between tumour and normal tissues.
Based on a computational reaction-diffusion model, Tzafriri [80] examined drug release from
intra-tumourally injected microspheres and drug transport in an idealized tumour model
accounting for binding in tumour interstitial fluid. Their simulation results suggested that
extracellular drug concentration was spatially heterogeneous in the initial period following drug
administration. It was concluded that increasing the duration of drug release might prolong the
2. LITERATURE REVIEW
39
exposure duration of anticancer drug but lower the plateau drug concentration. This implied that
intratumoural drug release could be optimized to improve treatment efficacy.
Baxter and Jain [88] developed a general theoretical framework to examine the effect of binding
on concentration profiles of macromolecules. A high binding affinity was predicted to be most
efficacious except for certain conditions of necrotic cores and low vascular permeability.
Doxorubicin was also found to bind extensively with proteins, mainly albumin, in both plasma
and interstitial fluid [55]. Due to the lack of in vivo experimental data, Eikenberry [90] assumed
that doxorubicin binding and unbinding in both plasma and interstitial fluid shared the same rates,
which were obtained based on experimental data [86]. Coupled with other drug transport
processes including transvascular and interstitial movement, drug binding was investigated in a
tumour core model. Their simulation results showed that most of the doxorubicin in extracellular
space existed in bound form.
2.3.4. Transport to Lymphatic System
The lack of a functional lymphatic system has been observed in solid tumours [49, 50]. The
limited existence of the lymphatic system can significantly reduce the clearance of large
molecular weight drugs and hence increase their retention in interstitial fluid. These are typically
liposome, nanoparticles and macromolecular drugs. On the other hand, without effective
drainage of interstitial fluid from the extracellular space, interstitial fluid pressure will increase,
which will in turn affect the drug exchange between blood and interstitial space.
Based on previous work on interstitial fluid pressure and drug convection, Baxter and Jain [50]
used a similar model to examine the effect of lymphatics on interstitial fluid pressure distribution
and concentration distribution of macromolecules (e.g. antibody). Their results suggested that
substances could be rapidly removed by lymphatics, if present in tumour, which would result in
much lower concentration levels.
2.3.5. Transport through Cell Membrane
After migrating in the interstitial space, drug can pass through cell membrane to enter tumour
cells. The main mechanisms for molecular transit across membranes are passive diffusion,
diffusion through channels, facilitated and active transport which involves the participation of
2. LITERATURE REVIEW
40
transmembrane proteins that bind a specific solute [46]. Drug transport through cell membrane is
a dynamics process, with influx and efflux occurring simultaneously [46, 90]. Dordal [98]
monitored cellular uptake of fluorescence-labelled doxorubicin in human lymphoma cells by a
flow cytometry assay with a rapid-injection system. The experimental data was then analysed by
a pharmacokinetic model to find the relation between drug concentration and cellular uptake.
Tumour cells can become resistant to anticancer drugs. One of the complex resistance
mechanisms is related to a 1.7×105
Da membrane protein named P-glycoprotein (P-gp). Being
insensitive to a wide variety of chemicals, P-gp is capable of actively transporting certain types
of anticancer drugs, including doxorubicin and vincristine. Luu [99] developed a mathematical
model to capture the efflux of doxorubicin resulting from drug mediated P-gp induction, and
they concluded that their mathematical model could be utilized to optimize dosing strategies in
order to minimize P-gp induction and thereby improving treatment efficacy.
El-Kareh and Secomb [55] included the dynamic process of cellular uptake and efflux in a
symmetric tumour model to investigate the influence of administration modes. Friedmen [100]
adopted a simplified model to fit cellular uptake rate data that showed saturation in both intra-
and extra-cellular concentrations. Kerr’s [101] experiments provided the experimental support to
this mathematical model. Cell uptake and efflux were analysed by human non-small cell lung
tumour cells in monolayer culture, and the relation between intracellular drug concentration and
cytotoxicity was obtained. A similar model was improved by Eikenberry [90] and then adopted
by Liu [102] to investigate the effects of ATP-binding cassette transporter-based acquired drug
resistance mechanisms at the cellular and tissue scale.
2.3.6. Cell Killing
After phagocytosis by tumour cells, drugs may destroy the genes that cause cells to divide or
interrupt the chemical process involved in cell division, and thereby killing cells. The rate of cell
killing has been found to be strongly dependent on extracellular and intracellular drug
concentrations, and several models have been proposed to describe this relationship.
The cytotoxicity of drug was firstly assumed to be a function of area under the extracellular
concentration-time curve (AUCe) [103]. Ozawa [104] derived a cell kill kinetics equation for cell
cycle phase-non-specific agents, and tested the equation by fitting to experimental results based
2. LITERATURE REVIEW
41
on Chinese hamster V79 cells. It was concluded that the cell killing action of cell cycle phase-
non-specific drugs could be described by a pharmacodynamics model with the concentration-
time product. Eichholtz-Wirth [105] treated Chinese hamster cells and Hela cells with
doxorubicin, and found the surviving cell fraction was proportional to the product of
extracellular drug concentration and exposure time. The experimental data also showed a non-
linear relation between extra- and intra-cellular concentrations.
Although there is evidence that drug can kill cells without entering them, cell death is believed to
be mainly dependent on intracellular concentration [1, 106]. Alternative models including
intracellular drug concentration and cellular pharmacodynamics have been proposed. A cellular
pharmacodynamics model for cisplatin was proposed by El-Kareh and Secomb [107] to relate
cell killing to the peak value of intracellular concentration over the entire treatment period. The
kinetics equations were modified for doxorubicin to reflect survival cell fraction [55, 90].
However, this model is incapable to tracking the dynamic variation of tumour cell density.
Given that chemotherapy is a dynamic process including cell proliferation, physical degradation
and cell killing caused by drugs, a dynamic model was suggested to evaluate the temporal
variation of cytotoxic effect of anticancer drugs [102]. Lankelma [108] proposed a
pharmacodynamics model for doxorubicin, where cell killing was described as a function of
intracellular concentration history. Drug accumulation history in cells was calculated using
cellular drug transport parameters derived from doxorubicin uptake and efflux measurements on
MCF-7 cells attached to culture dishes. Eliaz [109] developed a model taking into account the
cell proliferation rate, the cell killing rate, the average intracellular concentration and a lag time
for cell killing. The model was applied to free doxorubicin and doxorubicin encapsulated in
liposome targeted to melanoma cells in culture.
2.4. Summary
A detailed review of tumour characteristics, drug delivery methods and transport processes
indicates that all these three aspects depend on and interact with each other during anti-tumour
treatment. Although previous studies summarized in this chapter have addressed the complex
interplays among these factors to a certain extent, a systematic study that accounts for all these
physical and biological aspects is lacking. This is what the current project aims to achieve, and
2. LITERATURE REVIEW
42
the integrated mathematical model developed in this project will serve as a platform to compare
and optimise different strategies in chemotherapy.
3. Mathematical Models and Numerical
Procedures
In this chapter, a detailed description of mathematical models and their
implementations are given. The mathematical equations are described first; these
include equations governing the interstitial fluid flow in tumour and normal tissues,
mass transport and cellular uptake of anticancer agents, and heat transfer in response
to the application of HIFU for activation of drug release from thermo-sensitive
liposomes. Model parameters as well as relevant physical and transport properties are
then discussed. Finally, numerical methods to implement the mathematical models
are presented.
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
44
3.1. Mathematical Models
Drug properties are the key factors in anticancer treatments. In addition, drug transport also
depends on the interstitial fluid flow in tumour and its holding tissue, and the properties of
tumour cells have impacts on the cell killing as well. To include these relationships, the current
modelling platform consist of descriptions of interstitial fluid flow, transport for none-
encapsulated and thermo-sensitive liposome-mediated drugs, pharmacodynamics of tumour cells
and bioheat transfer model introduced by high intensity focus ultrasound heating for thermo-
sensitive liposome-mediated drugs only. Given the mentioned fact that the large size difference
exist between a tumour and its vasculature, a commonly adopted approach for study on drug
transport is treating the vasculature as a source term in the governing equations, without
considering its morphological features. The main assumptions are as follows: (1) the interstitial
fluid is incompressible, Newtonian fluid with homogeneous properties in tumour and
surrounding normal tissue, respectively; (2) all tumour cells are stationary, identical and alive
initially; (3) the simulation duration is significantly shorter than the timescale for tumour growth,
and thereby the physiological parameters and geometry are independent of time.
The mathematical models consist of the mass and momentum conservation equations for
interstitial fluid flow, mass transfer equations for the free and bound drug, transport equations for
liposome encapsulated drug, as well as equations describing the intracellular drug concentration
and pharmacodynamics. First, the interstitial fluid flow equations are solved to provide the basic
biomechanical environment for drug transport. This is followed by solution of the mass transfer
equations for drug transport, which is described schematically in Figure 3-1 (a) and (b) for direct
infusion, and thermo-sensitive liposome-mediated delivery of doxorubicin, respectively. Briefly,
the tumour region consists of three compartments: blood, extracellular space and tumour cells.
Within each compartment, letters F, B and L represent free, bound and liposome encapsulated
doxorubicin, respectively.
The dynamics process in drug delivery includes association/disassociation with protein at the rate
kas and kds respectively, drug exchange between blood and extracellular space, and influx/efflux
of drugs from extracellular space to tumour cells. The rate of cell killing is governed by a
pharmacodynamics model based on the predicted intracellular concentration of anticancer drugs.
In the case of thermo-sensitive liposome-mediated drug delivery, an additional equation
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
45
describing the transport of liposomes is needed. The detailed descriptions of each section of
mathematical models are listed below.
F
B
F
B
F
Blood Extracellular Space Tumour Cell
Cell
Killing
kas kds kas kds
influx
exfflux
clearance
exch
ange
(a)
F
B
L
F
B
L
Blood Extracellular Space
F
Tumour Cell
Cell
Killing
krel
kas kds kas kds
influx
exfflux
clearance
clearance
ex-
change
ex-
change
krel
(b)
Figure 3-1. Drug transport with (a) continuous infusion of drug in its free form, and (b) thermo-sensitive liposome-
mediated drug delivery
3.1.1. Interstitial Fluid Flow
The mass conservation equation for interstitial fluid is given by
lyv FF v (3-1)
where v is the velocity of the interstitial fluid, Fv and Fly are the source and sink terms,
accounting for the gain of interstitial fluid from blood vessels, and fluid absorption rate by the
lymphatics, respectively. Fv and Fly can be determined by Starling’s law
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
46
ivTivvv ppV
SKF (3-2)
where Kv is the hydraulic conductivity of the microvascular wall, S/V is the surface area of blood
vessels per unit volume of tumour tissue, pv and pi are the vascular and interstitial fluid pressures,
respectively, σT represents the average osmotic reflection coefficient for plasma protein, πv is the
osmotic pressure of the plasma, and πi is that of interstitial fluid.
The lymphatic drainage, Fly, is related to the pressure difference between the interstitial fluid and
lymphatics.
lyi
ly
lyly ppV
SKF (3-3)
where Kly is the hydraulic conductivity of the lymphatic wall, Sly/V is the surface area of
lymphatic vessels per unit volume of tumour tissue, and ply is the intra-lymphatic pressure.
The tumour and its surrounding tissues are treated as porous media, which can be justified as the
inter-capillary distance (33-98 µm [21, 110]) is typically 2-3 orders of magnitude smaller than
the length scale for drug transport. The corresponding momentum equation is given by
Fτvv
v
ip
t
(3-4)
where the stress tensor τ is expressed as
Ivvvτ )(3
2])([ T
(3-5)
where ρ and μ is the density and dynamic viscosity of interstitial fluid, respectively. I is the unit
tensor. The last term in equation (3-4), F, represents the Darcian resistance to fluid flow through
porous media and is given by
vvCvF 2
1W (3-6)
and W is a diagonal matrix with all diagonal elements calculated as
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
47
1 W (3-7)
where C is the prescribed matrix of the inertial loss term, and κ is the permeability of the
interstitial space. Since the interstitial fluid velocity is very slow (|v|<<1) [49], the inertial loss
term can be neglected when compared to the Darcian resistance. In addition, the interstitial fluid
is treated as incompressible with a constant viscosity. Hence, equation (3-6) can be reduced to
vF W (3-8)
3.1.2. Drug Transport
Drug transport is governed by the convection-diffusion equations for the free and bound drug in
the interstitial fluid. Free doxorubicin concentration in the interstitial fluid (Cfe) is described by
ifefefe
feSCDvC
t
C
2 (3-9)
where Dfe is the diffusion coefficient of free doxorubicin. The source term, Si, is the net rate of
doxorubicin gained from the surrounding environment, which is given by
ubvi SSSS (3-10)
here Sv, Sb and Su represent the net rate of doxorubicin gained from the blood/lymphatic vessels,
association/dissociation with bound doxorubicin-protein and influx/efflux from tumour cells,
respectively.
flfpv FFS (3-11)
where Ffp is the doxorubicin gained from the blood capillaries in tumour and normal tissues, and
Ffl is the loss of doxorubicin to the lymphatic vessels per unit volume of tissue. Using the pore
model [49, 50, 87, 89] for trans-capillary exchange, Ffp and Ffl can be expressed as
1
Pe1
fPe
f
eCC
V
SPCσFF fefpfefpdvfp (3-12)
felyfl CFF (3-13)
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
48
where Cfp is the concentration of doxorubicin in blood plasma, σd is the osmotic reflection
coefficient for the drug molecules, and Pfe is the permeability of vasculature wall to free
doxorubicin. Pef is the trans-capillary Peclet number defined as
V
SP
F
fe
dv
1Pef (3-14)
The net doxorubicin gained due to protein binding and cellular uptake is governed by
feabedb CkCkS (3-15)
ζDDS ccu (3-16)
where Dc is the tumour cell density, ka and kd are the doxorubicin-protein binding and
dissociation rates, respectively.
The convection-diffusion equation for bound doxorubicin (Cbe) is similar to equation (3-9)
except for the source terms.
bbebe
2
bebebe SFCDCt
C
v (3-17)
where Dbe is the diffusion coefficient of the bound doxorubicin-protein, and Fbe accounts for the
gain of bound doxorubicin from blood vessels, which is given by
1
Pe1
bPe
b
eCC
V
SPCFF bebpbebpdvbe (3-18)
where Pbe is the permeability of the vessel wall to bound doxorubicin, and Cbp is the bound
doxorubicin concentration in the plasma. The trans-capillary Peclet number is
V
SP
F
be
dv
1Peb (3-19)
Since only unbound doxorubicin can pass through the cell membrane [55, 90], the rate of cellular
uptake is a function of free doxorubicin concentration in the interstitial fluid.
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
49
t
Ci (3-20)
efe
fe
kC
CV
max (3-21)
ii
i
kC
CV
max (3-22)
where Ci is the intracellular doxorubicin concentration, and Vmax is the rate of trans-membrane
transport, ζ and ɛ are cellular uptake and efflux functions, ke and ki are constants obtained from
experimental data, and φ is the volume fraction of extracellular space.
3.1.3. Thermo-Sensitive Liposome-Mediated Drug Transport
Additional equations are required to describe the transport of liposome encapsulated doxorubicin,
in order to solve separately the encapsulated drug concentration and released drug concentration.
Equations for the transport of released drug, including those for free drug concentration in
plasma and interstitial fluid, bound drug concentration in plasma and interstitial fluid as well as
intracellular concentration, are described using the same equations given in the preceding section.
Similar to equation (3-9), the liposome encapsulated drug concentration in the interstitial fluid
(Cle) is governed by
llellele SCDCt
C
2v (3-23)
where Dl is the diffusion coefficient of liposomes. The source term, Sl is the net rate of liposomes
gained from the surrounding environment, which is given by
rlpl SSS (3-24)
Here, Slp represents the liposomes gained from plasma, and Sr represents released drug in the
interstitial fluid. Slp can be calculated by
lllplp FFS (3-25)
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
50
where Flp is the liposomes gained from the capillaries in tumour and normal tissues, and Fll is the
loss of liposomes through the lymphatic vessels per unit volume of tissue. Values for Flp and Fll
can be obtained using the pore model described by Equations (3-12) to (3-14) but with transport
properties corresponding to the thermo-sensitive liposome.
The amount of released doxorubicin in the interstitial fluid, Sr, is given by
lerelr CkS (3-26)
where krel is the release rate of doxorubicin from liposome.
The free doxorubicin concentration in blood plasma (Cfp) is governed by
bpdfpafpfpfp
B
Tr
fpCkCkCCLF
V
VS
t
C
(3-27)
where Ffp represents the free doxorubicin crossing the capillary wall into the interstitial fluid. VT
and VB are tumour volume and blood plasma volume, respectively. CLfp is the plasma clearance
of drug. ka and kd are the association and disassociation rates with proteins.
The bound doxorubicin concentration in blood plasma (Cbp) is given by
bpbpbe
B
Tbpdfpa
bpCCLF
V
VCkCk
t
C
(3-28)
where CLbp is the plasma clearance of bound doxorubicin.
The free doxorubicin concentration in interstitial fluid (Cfe) is described by Equation (3-9),
except that the source term contains an additional contribution from Sr, as described below.
rubvi SSSSS (3-29)
Expressions for the terms on the right hand side have been given previously (see equations (3-11)
to (3-16) and (3-26)).
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
51
3.1.4. Pharmacodynamics Model
The change of tumour cell density with time is described by a pharmacodynamics model as
given below [109].
2
50
maxcgcpc
i
ic DkDkDCEC
Cf
dt
dD
(3-30)
The first term on the right hand side represents the anticancer effect, where fmax is the cell-kill
rate constant and EC50 is the drug concentration producing 50% of fmax. kp and kg are cell
proliferation rate constant and physiological degradation rate, respectively. In this study, cell
proliferation and physiologic degradation are assumed to have reached equilibrium at the start of
each treatment.
3.1.5. Heat Transfer under High Intensity Focused Ultrasound (HIFU) Heating
In localized heating in anti-tumour therapy, both blood and tissue including the tumour and its
holding tissue are heated. Energy balances for tissue and blood are illustrated in Figure 3-2 (a)
and (b), respectively.
TISSUE
Qulltrasound
Qexchange
with blood
(a)
BLOOD
Qultrasound
Qexchange
with tumour
Qblood flow (Tb)Qblood flow (Tn)
(b)
Figure 3-2. Schematic representation of heat transfer under HIFU heating in (a) tumour and (b) blood
The temperature (T) of tissue and blood can be calculated by solving the following energy
balance equations [74, 111]:
tbtbbbttt
tt qTTwcTkt
Tc
2 (3-31)
bnbbbbbtbbbbbb
bb qTTwcTTwcTkt
Tc
2 (3-32)
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
52
where ρ is the density, c is the specific heat, k is the thermal conductivity, and w is the perfusion
rate of blood flow. q represents the ultrasound power deposition, and subscripts b and t represent
blood and tissue, respectively. Tn is the normal body temperature, which is assumed to be 37 ˚C.
In reality, heating of biological tissue by HIFU also contains the effect of acoustic non-linearity
[112, 113]. However, studies show that the non-linear wave propagation can be ignored if a focal
intensity is within the range of 100~1000 W/cm2 and the peak negative pressure in the range of
1~4 MPa [74, 114]. The ultrasound power deposition per unit volume is assumed to be
proportional to the local acoustic intensity Ia as follows
aIq 2 (3-33)
where α is the absorption coefficient, and the intensity Ia is defined as:
a
aa v
pI
2
2
(3-34)
va is the speed of ultrasound, and pa is the acoustic pressure which is represented by the real parts
of the following equation
aaca vikp (3-35)
here kac represents the acoustic velocity potential, and is given by
2a
ac vk (3-36)
ω is the angular frequency, and λ is the wave length of ultrasound.
For a spherical concave radiator S (Figure 3-3) with a uniform normal velocity, if the diameter or
width of a slightly curved radiator is large compared to the wave-length, the resulting acoustic
velocity potential, Ψ, in a restricted region can be written as [115]
2
0 011
1
2,
bdik
ddRResR acu
(3-37)
where u is the normal velocity of a slightly curved radiating surface, which is represented by
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
53
tieu
0u (3-38)
here u0 is a constant and t is time. In Equation (3-37), d is the distance from a point on the
radiator to a field point, which is given by
2
11
2 cossin2 zRRrRd (3-39)
2cos21 bhRz (3-40)
22
111 41 ARRr (3-41)
Referring to Figure 3-3, the ultrasound radiator is represented by the thick lines and the heated
point is Q. A and a are the radius of curvature and radius of the circular boundary, respectively, h
is the depth of the concave surface and b is the chord from the transducer centre O to the radiator
boundary.
Q1Q
r
C
Aa
b
h
R1
θ1 θ αα
x
O
R
d
S
Figure 3-3. Dimensions and coordinates of a spherical radiator [115].
3.2. Model Parameters
Since the growth of tumour and normal tissues is ignored, all the geometric and transport
parameters used in this study are assumed to be independent of time. Values adopted for these
are summarized in Tables 3-1, 3-2 and 3-3 at the end of this Chapter for parameters related to the
tissue, liposome and doxorubicin, respectively. Justifications for the choices of some of the
parameters are given below.
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
54
3.2.1. Tissue Related Transport Parameters
Blood vessel surface area to tissue volume ratio (S/V)
The ratio of surface area of blood vessel to tissue volume has a direct influence on the amount of
anticancer drug in the interstitial fluid. Its value depends strongly on the type of tissue and stage
of tumour growth [116]. Pappenheimer et al. measured this in normal tissues [117], while Baxter
and Jain [49] recommended using 70 cm-1
and 200 cm-1
for normal and tumour tissues,
respectively. The effect of this parameter on anticancer efficacy and its derivation from in vivo
MR images will be discussed in Chapter 5.
Tumour cell density (Dc)
The volume fraction of extracellular space in a tumour tissue ranges from 0.2 to 0.6 [118].
Assuming an average tumour cell diameter, Eikenberry [90] estimated that tumour cell density
ranged from 9.55×1015
to 1.53 ×1016
cells/m3. During chemotherapy treatment, tumour cell
density may change as a result of drug uptake, cell proliferation and physiological degradation.
3.2.2. Drug Related Transport Parameters
Vascular Permeability (P)
Vascular permeability coefficient measures the capacity of a blood vessel (often capillary in
tumour) wall to allow for the flow of substances in and out of the vasculature. The structure of
vessel wall and the molecular size of the transported substance are key determinants of
permeability [110].
(1) Free & bound doxorubicin
Estimates of this parameter reported in the literature usually correspond to ‘effective
permeability’, which is on the order of 10-7
m/s for albumin in both tumour and normal tissues.
For Sulforhodamine B (MW=559 Da), its permeability in normal tissue is 3.4×10-7
m/s [119].
Patent is another agent with a similar molecular weight (566 Da) to that of doxorubicin, and its
permeability is 3.95×10-7
m/s in normal tissue of male frog [120]. Eikenberry [90] assumed the
effective permeability of free doxorubicin in tumour to be in the range of 8.1×10-7
and 3.7×10-6
m/s, whereas Ribba et al. [121] used 3.0×10-6
m/s in their simulation work. Compared with
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
55
normal tissues, Gerlowski and Jain [122] found the vessel wall permeability to be 8 times higher
in tumour tissues. Using MW for interpolation, Goh [51] assumed the permeability of
doxorubicin in tumour tissue to be nearly 8 times higher than in normal tissue. In the present
study, values of permeability in tumour and normal tissues are assumed to be 3.0×10-6
m/s and
3.75×10-7
m/s, respectively. Wu et al. [119] measured the permeability of albumin
(corresponding to albumin-bound doxorubicin) in both tumour and granulating tissues, and found
the corresponding vasculature permeability to be 7.8±1.2×10-9
m/s and 2.5±0.8×10-9
m/s,
respectively. In this thesis, their average value is adopted as the permeability of albumin-bound
doxorubicin.
(2) Liposome encapsulated doxorubicin
The permeability of polyethylene glycol coated liposomes of 100 nm through tumour capillaries
was measured at 37 o
C by Yuan et al. [26] and Wu et al. [25] as 2.0×10-10
and 3.42±0.78×10-9
m/s, respectively. In normal granulation tissues permeability of the same liposomes was
8.0~9.0×10-10
m/s at the same temperature [25, 26].
Diffusion Coefficient (D)
Diffusion coefficient is the constant of proportionality between the molar flux owing to
molecular diffusion and the gradient in the concentration of the species, which is also known as
the driving force for diffusion.
(1) Free and bound doxorubicin
Diffusion coefficient is related to the molecular weight (MW) of the substance [15]. The
relationship between diffusion coefficient in water and molecular weight can be represented by
[15, 123]
d2a
d1w MWaD
(3-45)
This equation is used to fit the diffusion coefficients of common anticancer agents in water [124]
(Figure 3-4). Since doxorubicin has a MW of 544 Da, its diffusion coefficient in water at 37 oC
could be estimated as 3.83×10-10
m2/s. It has also been found that diffusion coefficient in
neoplastic tissue deviates from free diffusion in water [118]. For MWs in the range of 376 Da and
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
56
66,900 Da, the ratio of diffusion coefficient in tumour (D) to that in water (DW) was found to be
linearly related to MW
d4d3w
aMWaD
D (3-46)
Using equation (3-46), the ratio for doxorubicin is 0.8885 and the diffusion coefficient of
doxorubicin in tumour is 3.40×10-10
m2/s (Figure 3-4 (b)).
At 37 oC, the relationship between diffusion coefficient (D) and MW in normal tissues may be
obtained from in vitro experimental data [125].
690003210778.175.08
MWMWD (3-47)
Based on equation (3-47), the diffusion coefficient of doxorubicin in normal tissues is calculated
as 1.58×10-10
m2/s.
50 100 150 200 250 300 350 400 450 500 550 600 650
2
3
4
5
6
7
8
9
10
11
12
Dif
fusi
on C
oef
fici
eng i
n W
ater
(10
-10 m
2/s
)
Molecular Weight (Da)
Data from Literature
Fitting Curve
(a)
0 10000 20000 30000 40000 50000 60000 70000
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
DT/D
W
Molecular Weight (Da)
Data from Literature
Fitting Curve
(b)
Figure 3-4. Estimation of doxorubicin diffusion coefficient in tumour tissue. (a) The diffusion coefficient of
anticancer drugs in water as a function of molecular weight, using equation (3-45) [15, 123]. For the best fitted
curve for experimental data in reference [124], ad1 = 380.44, ad2 = 0.73. (b) The relation between diffusion
coefficient ratio (tumour to water) and molecular weight. For the best fitted line using equation (3-46) to fit
experimental data from reference [118], ad3 = -8.55948 10-6
, ad4 = 0.89317.
Doxorubicin can bind with high molecular weight proteins (e.g. albumin [90]) in the interstitial
fluid. In such a case, the diffusion coefficient of the bound component depends mainly on the
protein (e.g. albumin). The effective diffusion coefficient for albumin in VX2 carcinoma was
measured as 8.89×10-12
m2/s [90]. Since the bound component has a MW of approximately 69
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
57
kDa [51], its diffusion coefficient in normal tissue estimated by Equation (3-47) is 4.17×10-12
m2/s.
(2) Liposome encapsulated doxorubicin
For diffusion of spherical particles through liquid at low Reynolds number, the diffusion
coefficient in water can be calculated by solving the Einstein-Stokes equation [15].
rTk
D Bw 6
*
(3-48)
where kB is Boltzmann's constant, T* is the absolute temperature, μ is viscosity and r is the radius
of spherical particle. Assuming the liposomes are spherical with a uniform diameter of 100 nm,
their diffusion coefficient at 37oC is 5.82×10
-12 m
2/s.
Yuan found that for liposome particles of 90 nm in diameter, the ratio of permeability in tumour
and diffusion coefficient in water was around 20 m-1
[110], and the permeability was
2.0±1.6×10-10
m/s [26]. Based on these results, the diffusion coefficient is 1.0±0.8×10-11
m2/s.
Based on the Einstein-Stokes equation, the diffusion coefficient is inversely related to the
particle radius. Hence, the diffusion coefficient of liposomes with 100 nm diameter is estimated
to be 9.0±7.2×10-12
m2/s. The diffusion coefficients in tumour and normal tissues adopted in this
thesis are 9.0×10-12
m2/s and 5.8×10
-12 m
2/s, respectively.
Reflection Coefficient (σ)
The reflection coefficient determines the efficiency of the oncotic pressure gradient in driving
transport across the vascular wall. It is related to the sizes of drug and pores on the vasculature
wall [126]. For the same drug, this parameter may vary in different types of tissues [127]. Wolf
et al. [128] measured the reflection coefficient for albumin and found this to be 0.82±0.08. The
sizes of albumin and liposome are 3.5 nm and 100 nm, respectively. The reflection coefficient
for liposome is estimated to be greater than 0.90, hence it is assumed to be 0.95 in this study.
Drug Dose (Dd)
As a common anticancer drug, doxorubicin is widely used in chemotherapy to treat various types
of cancer, such as lymphoma, genitourinary, thyroid, and stomach cancer [3]. By interacting with
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
58
DNA in cells, doxorubicin can inhibit the process of DNA replication. Because of this
mechanism of action, high concentration of doxorubicin in normal tissues can cause serious
damage to healthy cells, known as side effects. In clinical therapy, the most serious toxicity is
life-threatening cardiomyopathy [129, 130], leading to heart failure. Side effects set a limit to the
lifetime dose a patient can receive, which is approximately 550 mg per unit body surface area [3].
The dose of doxorubicin in clinical use is related to the patient’s body surface area. In each
treatment cycle, the dose is between 50 to 75 mg/m2 [3].
Plasma Pharmacokinetics (Cv)
Doxorubicin concentration in blood plasma is modelled as an exponential decaying function of
time. The form of equations depends on the infusion mode.
For continuous infusion, a tri-exponential decay is assumed based on the plasma
pharmacokinetics of doxorubicin [55, 90].
Tteeα
Aee
α
Aee
α
A
T
DC
Tteα
Ae
α
Ae
α
A
T
DC
tαtαtαTαtαTα
d
dv
tαtαtα
d
dv
222211
321
111
111
3
3
2
2
1
1
3
3
2
2
1
1
(3-49)
where t refers to time, Dd is the dose of doxorubicin and Td is the infusion duration. A1, A2 and A3
are compartment parameters and α1, α2, α3 are compartment clearance rate.
For bolus injection, the terms involving α2, α3, A2 and A3 represent the compartments with much
slower elimination. The drug concentration is assumed to follow an exponential decay based on
the plasma pharmacokinetics of doxorubicin [55, 90].
tα
dv eADC 1
1
(3-50)
Free doxorubicin in plasma can easily bind with proteins, such as albumin. Greene et al. [86]
found that approximately 75±2.7 % doxorubicin is present in the bound form, and the percentage
binding is independent of doxorubicin and albumin concentrations. Hence for direct infusion, the
free (Cfp) and bound (Cbp) doxorubicin concentrations in plasma are given by:
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
59
(3-51)
where s is the percentage of bound doxorubicin, which is 25% in this study.
3.3. Implementations of Mathematical Models
Usually the location where anticancer agents are administrated into a patient’s body is remote
from the tumour, hence the influence of drug infusion on blood pressure in the tumour is ignored.
In addition, drug transport processes are assumed to have no influence on the interstitial fluid
pressure because of the small size of anticancer agents. In order to generate initial conditions for
the time-dependent numerical simulation, the interstitial fluid flow equations are firstly solved to
obtain a steady-state solution in the entire computational domain. Following this, the obtained
pressure and velocity values are applied at time zero for the simulation of drug transport and
cellular uptake in order to predict the treatment efficacy. The mathematical equations described
above are implemented into ANSYS Fluent, which is described in the following section.
3.3.1. ANSYS Fluent
ANSYS FLUENT is a widely used CFD code based on the finite volume method (FVM). The
physical domain is firstly divided into a large number of computational cells forming a
computational grid on which control volumes are defined where the variable of interest is located
in the centroid. Governing equations in the differential forms are then integrated over each of the
control volumes until the residual between two adjunct iterations is less than a given tolerance.
The conservation principle for the variables in each control volume is expressed by the resulting
discretised equation.
The employed solution algorithm is the SIMPLEC which uses a relationship between velocity
and pressure corrections to enforce mass conservation and to obtain the pressure field. The
momentum, mass and heat transfer equations are discretised using the second order UPWIND
scheme, in which quantities at cell faces are computed using a multi-dimensional linear
reconstruction approach [131]. In this approach, higher-order accuracy is achieved at cell faces
through a Taylor series expansion of the cell-centred solution about the cell centroid. For
vbp
vfp
CsC
sCC
1
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
60
transient problems, a fully implicit second order backward Euler scheme is adopted, which is
unconditionally stable with respect to time step size. The Gauss-Seidel method is adopted to
update values at nodal points by using the previously computed results as soon as they become
available.
3.3.2. User Defined Routines for Drug Transport
Mass transfer equations describing the transport of drugs are coded in C programming language
by using the User Defined Scalar (known as UDS), which is a function for user defined
programme be dynamically loaded with FLUENT.
These equations are solved based on the interstitial fluid field resolved by solving the continuity
and momentum equations. Drug transport equations are also discretised using the second order
UPWIND scheme, and the tolerance controlling the convergence is set to be 1×10-8
.
3.3.3. User Defined Routines for Heat Transfer
Heat transfer under HIFU heating in blood and tissue is also incorporated via UDS. These
equations are solved with the drug transport equations when simulating thermo-sensitive
liposome-mediated drug delivery. The second order UPWIND scheme is adopted for spatial
discretisation and the convergence tolerance is set as 1×10-10
.
3.3.4. Boundary Conditions
In order to solve the above equations numerically and to ensure physiological relevance of the
solutions, each of the dependent variable equations requires meaningful values at the boundary
of the computational domain. These values are known as boundary conditions. Typical
boundaries involved in the problems dealt with in this thesis include: interface between tumour
regions, tumour-normal tissue interface and external surface of normal tissue. The specification
of boundary conditions is problem-dependent, hence details will be given for each problem
described in individual chapters.
3.4. Summary
The numerical methods and tools described above were applied to a realistic prostate tumour
with a focus on understanding the transport steps of non-encapsulated drugs (Chapter 4), a liver
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
61
tumour with a focus on elucidating the effect of heterogeneous distribution of blood vessels and
3 prostate tumours with a focus on examining the effect of tumour size and microvascular
density (Chapter 5), and a realistic prostate tumour with a focus on investigating the thermo-
sensitive liposome-mediated drug delivery with hyperthermia induced by high intensity focused
ultrasound (Chapter 6). All computations were carried out on an HP Compaq 8100 Elite CMT
PC.
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
62
Table 3-1. Parameters for tumour and normal tissues.
Parameter Definition Unit Tumour Tissue Normal Tissue Reference
S/V Surface area of blood vessels per unit tissue volume m-1
20000 7000 [49-51, 88]
Kv Hydraulic conductivity of the micro-vascular wall m/Pa·s 2.10×10-11
2.70×10-12
[49-51, 88]
ρ Density of interstitial fluid kg/m3 1000 1000 [51]
µ Dynamic viscosity of interstitial fluid kg/m·s 0.00078 0.00078 [51]
1/κ Permeability of the interstitial space m-2
4.56×1016
2.21×1017
[49-51, 88]
pv Vascular fluid pressure Pa 2080 2080 [49-51, 88]
πv Osmotic pressure of the plasma Pa 2666 2666 [49-51, 88]
πi Osmotic pressure of interstitial fluid Pa 2000 1333 [49-51, 88]
σT Average osmotic reflection coefficient for plasma proteins 0.82 0.91 [49-51, 88]
KlySly/V Hydraulic conductivity of the lymphatic wall times surface
area of lymphatic vessels per unit volume of tumour tissue
(Pa·s)-1
0 4.17×10-7
[51]
ply Intra-lymphatic pressure Pa 0 0 [51]
Dc Cell density 105cell/m
3 1×10
10 - [90]
VT Total tumour volume m3
5×10-5
3×10-4
[55]
VB Total blood volume in body m3 5×10
-2 5×10
-2 [55]
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
63
Table 3-2. Parameters for liposome
Parameter Definition Unit in Tumour in Normal Tissue Reference
Pl Liposome permeability of vasculature wall m/s 3.42×10-9
8.50×10-10
[25, 26]
D Liposome diffusion coefficient m2/s 9.0×10
-12 5.8×10
-12 [25, 110]
σl Reflection coefficient for liposome 0.95 1.0 -
CLlp Plasma clearance in tumour
s-1
2.228×10-4
2.228×10-4
[62]
3. MATHEMATICAL MODELS AND NUMERICAL PROCEDURES
64
Table 3-3. Parameters for doxorubicin.
Parameter Definition Unit Free Doxorubicin Bound Doxorubicin Reference
Ptumour Permeability of vasculature wall in tumour m/s 3.00×10-6
7.80×10-9
[51, 119]
Pnormal Permeability of vasculature wall in normal tissue m/s 3.75×10-7
2.50×10-9
[51, 119]
Dtumour Diffusion coefficient in interstitial fluid of tumour m2/s 3.40×10
-10 8.89×10
-12 [15, 51, 118, 123-125]
Dnormal Diffusion coefficient in interstitial fluid of normal tissue m2/s 1.58×10
-10 4.17×10
-12 [15, 51, 118, 123-125]
σd Osmotic reflection coefficient 0.15 0.82 [51, 128]
ka Doxorubicin-protein binding rate s-1
0.833 - [90]
kd Doxorubicin-protein dissociation rate s-1
- 0.278 [90]
φ Tumour fraction extracellular space 0.4 [90]
Vmax Rate of trans-membrane transport kg/105cells s 4.67×10
-15 - [90, 101]
ke Michaelis constant for transmembrane transport kg/m3
2.19×10-4
- [90, 101]
ki Michaelis constant for transmembrane transport kg/105cells 1.37×10
-12 - [90, 101]
fmax Cell-kill rate constant s-1
1.67×10-5
- [109]
EC50 Drug concentration producing 50% of fmax kg/105cells 5×10
-13 - [109]
A1 parameter for pharmacokinetic model m-3
74.6 74.6 [85, 90]
A2 parameter for pharmacokinetic model m-3
2.49 2.49 [85, 90]
A3 parameter for pharmacokinetic model m-3
0.552 0.552 [85, 90]
α1 compartment clearance rate s-1
2.43×10-3
2.43×10-3
[85, 90]
α2 compartment clearance rate s-1
2.83×10-4
2.83×10-4
[85, 90]
α3 compartment clearance rate model s-1
1.18×10-5
1.18×10-5
[85, 90]
kp Cell proliferation rate s-1
3.0×10-6
- [102]
kg Cell physiologic degradation rate s-1
3.0×10-16
- [102]
CLtumour Plasma clearance in tumour s-1
2.43×10-3
0 [85, 114, 132, 133]
CLnormal Plasma clearance in normal tissue s-1
2.43×10-3
0 [85, 114, 132, 133]
4. Non-Encapsulated Drug Delivery to Solid
Tumour
Intravenous administration of drugs in their free form is a traditional drug delivery
method adopted in clinical treatments without specific targeting. Injected drugs
undergo multiple transport processes in systemic blood stream, interstitial space and
tumour cells. The role of drug transport in treatment efficacy is complicated because
of nonlinear interactions between the drug and the microenvironment of tumour and
surrounding normal tissues. An appropriate mathematical model is therefore needed
to gain detailed understanding of the drug transport processes involved and how drug
concentrations and its therapeutic effect may be influenced by infusion modes.
In this chapter, a coupled pharmacokinetics and drug transport model is applied to a
real tumour geometry reconstructed from magnetic resonance image (MRI) to
examine the effects of various administration modes and clinic dose levels. This is
followed by comparisons of drug concentrations and survival fraction of tumour cells
between 2-D and 3-D geometry models.
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
66
4.1. Model Description
The mathematic model described in Chapter 3 for free form drug administration is adopted,
which incorporates the key physical and biochemical processes, including time-dependent
plasma clearance, drug transport through the blood and lymphatic vessels, extracellular drug
transport (convection and diffusion), drug binding with proteins, lymphatic drainage, interactions
with the surrounding normal tissue and drug uptake by tumour cells. Anticancer efficacy is
evaluated based on the percentage of survival tumour cells by directly solving the
pharmacodynamics equation using the predicted intracellular drug concentration.
4.1.1. Model Geometry
The geometry of a prostate tumour is reconstructed from images acquired from a patient using a
3.0-Tesla MR scanner (DISCOVERY MR750, GE, Schenectady, New York, USA). Multislice
anatomical images of the prostate were acquired in three orthogonal planes with echo-planer (EP)
sequence, with each image comprising 256 by 256 pixels. Other imaging parameters are given in
Table 4-1.
Table 4-1. MR imaging parameters
Parameter
(unit)
Pixel Size
(mm)
Field of View
(cm)
Slice Thickness
(mm)
Repetition Time
(ms)
Echo Time
(ms)
1.250 32.0 7.000 4000 84.4
An example of the MR images is shown in Figure 4-1 (a) with the tumour region and its
surrounding normal prostatic tissues. Transverse images are processed using image analysis
software Mimics (Materialise HQ, Leuven, Belgium), and the tumour is segmented from its
surrounding normal tissue based on signal intensity values. The resulting smoothed surfaces of
the tumour and normal tissues are imported into ANSYS ICEM CFD to generate computational
mesh for the entire volume.
The tumour shown in Figure 4-1 (b) is located at the corner of normal tissue with a dimension of
47 mm (maximum width) by 38 mm (maximum depth) in 2-D model. The final mesh consists of
64,966 triangular elements which have been tested to produce grid independent solutions.
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
67
(a)(a)
(b)
Tumour
Normal Tissue
(b)
Figure 4-1. Model geometry: (a) MR image of the prostate tumour (in red) and its surrounding tissue (in pule blue);
(b) the reconstructed 2-D geometry.
4.1.2. Model Parameters
The parameters describing general properties of doxorubicin, tumour and normal tissue have
been given in Tables 3-1 and 3-3 in the preceding chapter. Infusion durations and doses
simulated in this study are described below.
(1) Infusion Duration
Infusion duration is a key parameter in determining the pharmacokinetics of drug in blood.
Previous experiments found that prolonged infusion, such as over 48 hours or 96 hours, did not
contribution to better treatment efficacy [53, 54]. In a numerical study by [55], a 2-hour infusion
was adopted without the consideration of cell proliferation and death as a dynamic process.
Therefore, bolus injection and continuous infusion with infusion duration of 1-hour, 2-hour and
3-hour respectively, are studied.
(2) Infusion Dose
Limited by the serious side effect caused by doxorubicin, the lifetime dose a patient can receive
is approximately 550 mg per unit body surface area. For each treatment cycle, the dose is
between 50 to 75 mg/m2 [3], depending on the body surface area of a patient.
Body surface area (BSA) is determined from the following formula [51]
(4-1) 2
730
m731kg70
kginWeightBSA .
.
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
68
For a 70 kg patient, dosage of doxorubicin is in the range of 86.5~129.75 mg. Three dose levels,
50, 60 and 75 mg/m2 are examined in this study.
(3) Pharmacokinetics
Doxorubicin concentration in blood plasma is usually described by an exponential function over
time. Better fit to the experimental data was found by using a 3-compartment model.
For bolus injection and continuous infusion in a single administration, the plasma concentration
of doxorubicin can be described by Equation (3-49) and (3-50), respectively. The plasma
concentration in multiple administrations under bolus injection and continuous infusion are given
below by Equation (4-2) and (4-3), respectively [134].
(4-2)
(4-3)
where Dd is the dose of doxorubicin and Td is the infusion duration. A1, A2 and A3 are
compartment parameters and α1, α2, α3 are compartment clearance rate. t* is the time after n
doses (i = 1,2,…,n) given at time tDi.
To account for the high binding rate for doxorubicin with proteins, such as albumin in blood
plasma, and for the fact that the percentage of bound doxorubicin is independent of plasma
concentration of doxorubicin and albumin, the free (Cfp) and bound (Cbp) doxorubicin in blood
are governed by Equation (3-51), where s is the percentage of bound doxorubicin, which is 0.25
in this study [86].
n
i
tt
dviD
*
eADC1
1
1
n
i
dD
*
TttT
TttTTttT
di
di
dD
*
tt
tt
tt
dn
dnn
i
TttT
TttT
TttT
di
di
v
Ttt
eeA
eeA
eeA
T
D
Ttt
eA
eA
eA
T
D
eeA
eeA
eeA
T
D
C
n
diiD*
di
diiD*
didiiD*
i
n
nD*
nD*
nD*
diiD*
di
diiD*
di
diiD*
di
1
3
3
2
2
1
1
3
3
2
2
1
1
1
1
3
3
2
2
1
1
33
2211
3
2
1
33
22
11
1
11
1
1
1
1
1
1
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
69
4.1.3. Numerical Methods
The mathematical models described above are implemented in ANSYS-Fluent, which is a finite
volume based computational fluid dynamics (CFD) code (ANSYS Inc., Canonsburg, USA). The
momentum and drug transport equations are discretised using the second order UPWIND scheme,
and the SIMPLEC algorithm is employed for pressure-velocity coupling. The Gauss-Seidel
smoothing method is used to update values at nodal points after each iteration step. Convergence
is controlled by setting residual tolerances of the momentum equation and the drug transport
equations to be 1×10-5
and 1×10-8
, respectively.
In order to generate initial conditions for the transient simulation, the interstitial fluid flow
equations are firstly solved to obtain a steady-state solution in the entire computational domain.
Following this, the obtained pressure and velocity values are applied at time zero for the
simulation of drug transport and cellular uptake (shown in Figure 4-2). The second order implicit
backward Euler scheme is used for temporal discretisation, and a fixed time step size of 10
seconds is chosen. This time step is deemed sufficiently fine based on a time step sensitivity test.
The initial doxorubicin concentrations are assumed to be zero in both tumour and the
surrounding normal tissue.
Solving continuity and momentum
equations to obtain steady-state
solution of interstitial fluid field
Solving mass transfer equations to
obtain time-dependent drug
concentration and tumour cell
denisty
IFP IFV
Figure 4-2. Numerical Procedures.
There are two boundary surfaces in this model: an internal boundary between the tumour tissue
and normal tissue, and the outer surface of the normal tissue. At the internal boundary,
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
70
conditions of continuity of the interstitial pressure and fluid flux are applied. At the outer surface,
a constant relative pressure of 0 Pa and zero flux of drug are assumed.
4.2. Results
The analysis of results obtained from the mathematical model is carried out. The interstitial fluid
field is firstly examined to provide the microenvironment for drug transport and cell killing. This
is followed by comparisons of administration modes and doses under single and multiple
administration treatments in a clinical range. Finally, computational results obtained from the 2-
D and 3-D models are compared.
4.2.1. Interstitial Fluid Field
The interstitial fluid field in tumour and normal tissues plays an important role in drug delivery.
This is because on the one hand, the interstitial fluid velocity and pressure determines the drug
migration in extracellular space by convention; on the other hand, transvascular pressure gradient
influences the drug exchange between microvessels and the interstitial space of tumour. By
solving Equations (3-1)~(3-8) described in Chapter 3 for a vascular pressure of 2080 Pa [49-51,
89], the spatial distribution of interstitial fluid pressure (IFP) is presented in Figure 4-3. There is
a uniform IFP in the tumour and normal tissue. Higher IFP is found in the tumour compared with
its surrounding normal tissue, and there is a large gradient in a thin layer at the interface between
the two regions.
Figure 4- 3. Interstitial fluid pressure in tumour and surrounding normal tissue.
The spatial mean IFP is 1533.88 Pa in the tumour region and 40 Pa in normal tissue. These
results agree with experimental studies which suggested that IFP in tumour was in the range of
586.67 Pa to 4200 Pa [48], and in normal tissue -400 Pa to 800 Pa [135]. This difference in IFP
between tumour and normal tissue can be partially attributed to the special characteristics of
1400
1173
947
720
493
266
40
IFP (Pa)
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
71
tumour vasculature. On the one hand, vasculature surface is enlarged because blood vessels are
tortuous, elongated, and often dilated in tumour growth. The hydraulic conductivity is increased
by 10-fold as a result of large pores on tumour vasculature wall [49]. This leads to a much easier
fluid exchange across the blood vessel wall. On the other hand, the lack of lymphatic vessels
results in less fluid being drained out of the interstitial space, causing build-up of pressure in
tumour interstitial space. The IFP distribution shown in Figure 4-3 suggests that pressure-
induced convection of interstitial fluid mainly occurs within a thin layer at the interface between
tumour and its surrounding normal tissue.
In theory, high pressure in tumour poses a barrier for drug transport from blood to interstitial
space. However, diffusion driven by concentration gradient and vasculature surface area also
contributes to drug exchange between blood and interstitial space. Previous studies demonstrated
that diffusion was a more important mechanism in determining drug transport across the
vasculature wall because of the ineffective convection owing to the high IFP [51].
4.2.2. Single Administration Mode
In this series of simulations, the full dose of doxorubicin is administrated at the beginning of
each treatment. The influences of two controllable factors, infusion duration and dose level, are
examined over a period of 9 hours following drug administration.
4.2.2.1. Infusion Duration
The cytotoxic effect of doxorubicin on tumour cells is evaluated for bolus injection and
continuous infusions with different infusion durations. Given that doxorubicin is carried by the
blood stream into the tumour and normal tissue, its concentration in blood should be examined
when evaluating its anticancer effectiveness. The time course of doxorubicin blood concentration,
for a total dose of 50 mg/m2 administered by different administration modes, is compared in
Figure 4-4. Because all the drugs are administrated into blood stream in a very short period,
which can be ignored in comparison with the treatment time window, drug concentration in
blood under bolus injection reaches its peak at the very beginning of treatment. Without
continuous supply, the concentration falls rapidly to a low level. Continuous infusion leads to
different results. Since the total dose of drugs are administrated continuously in the infusion
duration, drug concentration in blood increases gradually in this period. At the end of the
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
72
infusion duration, drug concentration begins to decrease. Therefore, the timing of peak drug
concentration varies according to the infusion duration.
0 1 2 3 4 5 6 7 8 9
0.0
2.0x10-3
4.0x10-3
6.0x10-3
8.0x10-3
1.0x10-2
1.2x10-2
Doxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n (
kg/m
3)
Time (hr)
bolus injection
1-hour infusion
2-hour infusion
3-hour infusion
4 5 6 7 8 92.0x10
-5
4.0x10-5
6.0x10-5
8.0x10-5
Figure 4-4. Doxorubicin concentration in plasma as a function of time after start of treatment, for bolus injection and
continuous infusion of various indicated durations (dose = 50 mg/m2).
Comparison of peak values in Figure 4-4 indicates that prolonging the infusion duration results
in low peak doxorubicin blood concentration. This is because doxorubicin is being continuously
cleared up after administration. Since high doxorubicin concentration in blood is associated with
increased risk of cardiotoxicity, bolus injection may not be a preferred option while continuous
infusion with longer duration may help reduce this risk. Comparison also shows that although
fast infusion leads to higher concentrations at the beginning of the treatment, doxorubicin
concentration in blood falls much faster after the infusion ends.
Free and bound doxorubicin extracellular concentrations in tumour and its surrounding normal
tissue are shown in Figure 4-5 and 4-6, respectively. Regardless of the injection mode, both free
and bound doxorubicin concentrations increase rapidly during the initial period following drug
administration. Doxorubicin concentrations in tumour interstitial space increase faster and reach
a much higher peak with bolus injection. The peak value for continuous infusion decreases with
the increase in infusion duration. Although doxorubicin concentration drops to a low level after
infusions end for all modes of administration, continuous infusions are able to maintain a slightly
higher concentration than bolus injection. Comparing doxorubicin extracellular concentrations in
Figures 4-5 and 4-6 with the concentration in blood in Figure 4-4, the concentration curves have
identical shapes for a given administration mode. This means that drug concentration in blood
has a direct influence on the extracellular concentration for both free and bound doxorubicin.
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
73
0 1 2 3 4 5 6 7 8 9
0.0
5.0x10-4
1.0x10-3
1.5x10-3
2.0x10-3
2.5x10-3
3.0x10-3
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
bolus injection
1-hour infusion
2-hour infusion
3-hour infusion
4 5 6 7 8 90.0
8.0x10-6
1.6x10-5
2.4x10-5
3.2x10-5(a)
0 1 2 3 4 5 6 7 8 9
0.0
1.0x10-3
2.0x10-3
3.0x10-3
4.0x10-3
5.0x10-3
6.0x10-3
7.0x10-3
8.0x10-3
9.0x10-3
Boudn D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
bolus injection
1-hour infusion
2-hour infusion
3-hour infusion
4 5 6 7 8 90.0
3.0x10-5
6.0x10-5
9.0x10-5
1.2x10-4
(b)
Figure 4-5. Spatial mean (a) free and (b) bound doxorubicin extracellular concentration in tumour as a function of
time under bolus injection and different infusion durations (dose = 50 mg/m2).
Because binding with protein is a dynamics process, both free and bound doxorubicin
concentrations vary with time. Comparing Figure 4-5 (a) and (b), it is clear that the time course
of free and bound concentrations shares a similar trend, but bound doxorubicin concentration is
approximately 3 times of the free doxorubicin concentration, indicating that most doxorubicin in
the interstitial fluid is in bound form. Since only free form doxorubicin can be taken up by
tumour cells [55, 90], the high binding rate is expected to reduce the intracellular concentration
of doxorubicin.
0 1 2 3 4 5 6 7 8 9
0.0
1.0x10-4
2.0x10-4
3.0x10-4
4.0x10-4
5.0x10-4
6.0x10-4
7.0x10-4
8.0x10-4
9.0x10-4
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
bolus injection
1-hour infusion
2-hour infusion
3-hour infusion
(a)
0 1 2 3 4 5 6 7 8 9
0.0
5.0x10-4
1.0x10-3
1.5x10-3
2.0x10-3
2.5x10-3
3.0x10-3
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
bolus injection
1-hour infusion
2-hour infusion
3-hour infusion
(b)
Figure 4-6. Spatial mean (a) free and (b) bound doxorubicin extracellular concentration in normal tissue as a
function of time under bolus injection and different infusion durations. (dose = 50 mg/m2).
Figure 4-7 shows the spatial distribution of free doxorubicin extracellular concentration in
tumour and its surrounding normal tissue at 0.5 hour under 2-hour continuous infusion. There is
a difference in drug concentration between tumour and surrounding normal tissue. Since the
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
74
properties of doxorubicin and tissues are assumed to be uniform in each region, the
corresponding distribution is also uniform except near the tumour boundary where a large
concentration gradient exists. This thin layer of steep concentration gradient represents the
exchange of drug between tumour and normal tissue.
3.58×10-5
3.38×10-5
3.19×10-5
3.00×10-5
2.79×10-5
2.60×10-5
2.40×10-5
Cf (kg/m3)
Figure 4-7. Spatial distribution of extracellular concentration of free doxorubicin in tumour and normal tissues (2-
hour infusion, time = 0.5 hour).
As has already been shown in Figure 4-3, the interstitial fluid pressure in tumour is much higher
than in normal tissue; this pressure difference drives interstitial fluid flow from the tumour to
normal tissue, and doxorubicin is also carried by this flux. Moreover, doxorubicin diffuses from
tumour where its concentration is high compared to the surrounding tissue because of the large
concentration gradient in the interstitial space. The transport from tumour to normal tissue needs
to be minimised in order to maintain a high level of doxorubicin concentration in tumour.
0 1 2 3 4 5 6 7 8 9
0.0
0.5
1.0
1.5
2.0
Doxoru
bic
in I
ntr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(ng/1
05ce
lls)
Time (hr)
bolus injection
1-hour infusion
2-hour infusion
3-hour infusion
Figure 4-8. Doxorubicin intracellular concentration as a function of time, for bolus injection and continuous
infusions of various durations (dose = 50 mg/m2).
Figure 4-8 presents the intracellular doxorubicin concentration in tumour under bolus injection
and continuous infusions with various infusion durations. The intracellular concentration
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
75
displays a sharp rise at the beginning until it reaches a peak, and then decreases. During the
initial phase, the rate of increase in doxorubicin concentration slows down with increase of
infusion duration. Compared to continuous infusions, bolus injection leads to a higher peak and a
faster reduction in intracellular concentration which also remains at a much lower level after
cessation of drug administration.
For continuous infusions, longer infusion durations tend to slow down the increase in
intracellular concentration at the initial phase of treatment, and produce a lower peak. Here, the
highest intracellular concentration is found for a 1-hour continuous infusion. Moreover, for
continuous infusion with different durations, intracellular concentrations reach a very similar
level after drug infusions terminate.
0 1 2 3 4 5 6 7 8 9
76
80
84
88
92
96
100
Fra
ctio
n o
f S
urv
ival
Cel
l (%
)
Time (hr)
bolus injection
1-hour infusion
2-hour infusion
3-hour infusion
Figure 4-9. Predicated percentage tumour cell survival under bolus injection and continuous infusions of various
durations (dose = 50 mg/m2).
Figure 4-9 shows the percentage of survival tumour cells by applying the pharmacodynamics
model described by Equation (3-30). As can be observed, the cytotoxic effect of bolus injection
on tumour cells is significantly lower than that of continuous infusions. This is because of the
rapid clearance of doxorubicin from blood (shown in Figure 4-4), resulting in a dramatic
reduction in extracellular drug concentration which is reduced to approximately zero after 5~6
hours (shown in Figure 4-5). Without enough doxorubicin accumulating in tumour cells, the cell
killing rate and physical degradation rate become slower than cell proliferation rate, and hence
tumour cell density begins to increase after reaching the minimum value at around 5 hours after
the cessation of drug administration. On the contrary, the extracellular drug concentration with
continuous infusion stays at a low but non-zero level for a longer period to supply sufficient
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
76
doxorubicin for cell killing. This is the reason for the continuous reduction in survival fraction of
tumour cells with continuous infusion in the simulation time window.
However, the effect of infusion duration seems to be insignificant for the limited range of
infusion durations examined, with a difference of up to 2% between the slower (3 hours) and
faster infusion (1 hour). The best therapeutic efficacy is not achieved with the infusion duration
that leads to the highest intracellular concentration. Results show that 1-hour infusion leads to
high cell killing rate at the beginning of treatment because of high intracellular concentration
(shown in Figure 4-8). However, the killing rate slows down due to a rapid reduction in
intracellular concentration. This low level of intracellular concentration sustains to the end of the
simulation time window, and the therapeutic efficacy is the worst among the continuous infusion
modes examined. On the contrary, although a 3-hour infusion results in a slower increase and
lower peak, as shown in Figure 4-8, the intracellular concentration is at a higher level after peak
time and the cytotoxic effect is better than others. This indicates that the final treatment outcome
is more dependent on the intracellular concentration over the entire exposure time rather than its
peak value.
4.2.2.2. Effect of Dose Level
Another controllable parameter in chemotherapy is the dose level of anticancer drugs. In current
clinical use, doxorubicin dose ranges from 50 mg/m2 to 75 mg/m
2 for a 70kg patient, hence the
effect of dose on drug effectiveness is examined within this range.
Comparison is made between different doses (50, 60 and 75 mg/m2) under a 2-hour continuous
infusion, simulating treatment doses used for a 70 kg patient. Blood concentration of free
doxorubicin presented in Figure 4-10 (a) shows that increasing the dose results in linear elevation
of drug concentration. Since following the same pharmacokinetics, blood concentration of bound
doxorubicin is believed to share the same feature. As shown in Figure 4-10 (b) and (c),
extracellular concentration of doxorubicin in both free and bound forms are increased because of
the close relationship between concentration in blood and in interstitial space. As a consequence,
more drugs accumulate in tumour cells under a high dose administration.
Results shown in Figure 4-10 (d) show that intracellular drug concentration is not proportional to
the administrated dose level. In summary, doxorubicin concentrations in all regions increase with
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
77
the dose level, since a high dose increases the amount of drug delivered to both the tumour and
normal tissues. In addition, the high concentration in blood may also increase the risk of heart
failure.
0 1 2 3 4 5 6 7 8 9
0.0
1.0x10-4
2.0x10-4
3.0x10-4
4.0x10-4
5.0x10-4
6.0x10-4
7.0x10-4
8.0x10-4
Doxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n (
kg/m
3)
Time (hr)
75 mg/m2
60 mg/m2
50 mg/m2
(a)
0 1 2 3 4 5 6 7 8 9
0.0
2.0x10-5
4.0x10-5
6.0x10-5
8.0x10-5
1.0x10-4
1.2x10-4
1.4x10-4
1.6x10-4
1.8x10-4
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
75 mg/m2
60 mg/m2
50 mg/m2
(b)
0 1 2 3 4 5 6 7 8 9
0.0
1.0x10-4
2.0x10-4
3.0x10-4
4.0x10-4
5.0x10-4
6.0x10-4
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
75 mg/m2
60 mg/m2
50 mg/m2
(c)
0 1 2 3 4 5 6 7 8 9
0.0
0.5
1.0
1.5
2.0
2.5D
oxoru
bic
in I
ntr
acel
lula
r C
once
ntr
atio
n
(ng/1
05ce
lls)
Time (hr)
75 mg/m2
60 mg/m2
50 mg/m2
(d)
Figure 4-10. Spatial mean doxorubicin concentration in tumour as a function of time under 2-hour infusion
duration.(a) doxorubicin blood concentration, (b) free doxorubicin extracellular concentration, (c) bound
doxorubicin extracellular concentration, and (d) intracellular concentration.
The percentage of survival tumour cells is compared in Figure 4-11 by solving the
pharmacodynamics equation based on the predicted intracellular concentration. The results show
that the cytotoxic effect of doxorubicin on tumour cells increases with the dose level, with the
difference between low (50 mg/m2) and high doses (75 mg/m
2) being 2.5% at 9-hour following
drug administration. Compared to the administrated dose level, this improvement in treatment
outcome is not obvious.
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
78
0 1 2 3 4 5 6 7 8 9
75
80
85
90
95
100
Fra
ctio
n o
f S
urv
ival
Cel
l (%
)
Time (hr)
75 mg/m2
60 mg/m2
50 mg/m2
Figure 4-11. Predicated percentage of tumour cell survival under various doses.
The predicated free and bound doxorubicin concentration in the interstitial space of normal tissue
with different drug dosages is compared in Figure 4-12. Results show that increasing the dose
results in elevation of extracellular concentration of doxorubicin in both free and bound forms.
This indicates that the risk of cell killing in normal tissue may also be increased at higher dose.
0 1 2 3 4 5 6 7 8 9
0.0
4.0x10-5
8.0x10-5
1.2x10-4
1.6x10-4
2.0x10-4
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
75 mg/m2
60 mg/m2
50 mg/m2
(a)
0 1 2 3 4 5 6 7 8 9
0.0
1.0x10-4
2.0x10-4
3.0x10-4
4.0x10-4
5.0x10-4
6.0x10-4
75 mg/m2
60 mg/m2
50 mg/m2
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
(b)
Figure 4-12. Spatial mean doxorubicin concentration in normal tissue as a function of time under 2-hour infusion
duration (for a 70 kg patient). (a) free doxorubicin extracellular concentration, (b) bound doxorubicin extracellular
concentration.
4.2.3. Multiple Administration Mode
Numerical studies are carried out under continuous infusion and bolus injection with multiple
administrations for 2, 4 and 8 times within the first 48 hours of treatments with intervals of 24,
12 and 6 hours, respectively. The infusion duration is set as 2-hour for each of continuous
infusion, and the simulation is performed for 96 hours.
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
79
4.2.3.1. Continuous Infusion
Figure 4-13 shows the predicted time course of plasma doxorubicin concentration, for a total
dose of 50 mg/m2 administrated by single and multiple 2-hour continuous infusions. Regardless
of the administration mode, the drug concentration in blood reaches a peak after the end of each
administration. Results show that multiple-infusion can significantly reduce the peak plasma
concentration, and this reduction effect increases with the number of infusion. Administrating the
same dose of doxorubicin 8-times gives a 20% lower peak concentration compared with a single
administration. Hence multiple-infusion is likely to reduce the risk of heart electrophysiology
dysfunction and muscle damage caused by doxorubicin. However, quantitative comparison of
blood concentration after administration shows that multiple-infusion can also lead to a slightly
high level of drug concentration in plasma over time.
Figure 4-13. Doxorubicin intravascular concentration under single and multiple continuous infusions as a function of
time (total dose = 50 mg/m2).
Figure 4-14 shows the time course of free and bound doxorubicin concentration in the interstitial
fluid for different infusion modes. During each infusion period, all forms of doxorubicin begin to
accumulate in the interstitial space since doxorubicin is continuously administrated into the
circulation system. Drug concentration reaches a peak value at the end of each infusion, and then
falls to a lower level as time proceeds. Results show that free and bound doxorubicin
concentrations share the same trend for a specific administration mode. Multiple-infusion can
effectively reduce the peak extracellular concentration and lead to a slightly higher concentration
level over time.
0 8 16 24 32 40 48 56 64 72 80 88 96
0.0
1.0x10-4
2.0x10-4
3.0x10-4
4.0x10-4
5.0x10-4
Doxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
(kg/m
3)
Time (hr)
single_infusion
twice_infusion
4 times_infusion
8 times_infusion
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
80
0 12 24 36 48 60 72 84 96
0.0
2.0x10-5
4.0x10-5
6.0x10-5
8.0x10-5
1.0x10-4
1.2x10-4
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
single_infusion
twice_infusion
4_times_infusion
8_times_infusion
(a)
0 12 24 36 48 60 72 84 96
0.0
5.0x10-5
1.0x10-4
1.5x10-4
2.0x10-4
2.5x10-4
3.0x10-4
3.5x10-4
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
single_infusion
twice_infusion
4_times_infusion
8_times_infusion
(b)
Figure 4-14. Spatial mean doxorubicin extracellular concentration as a function of time under single and multiple
continuous infusion. (a) free doxorubicin, and (b) bound doxorubicin (total dose = 50 mg/m2).
Intracellular doxorubicin concentrations in tumour with signal and multiple-infusion modes are
compared in Figure 4-15. It is clear that single infusion gives the highest peak intracellular
concentration, while multiple-continuous-infusions have lower but multiple peaks which help to
keep a sustained level of intracellular concentration over the entire treatment period.
Figure 4-15. Time course of doxorubicin intracellular concentration under single and multiple continuous infusions.
(total dose = 50 mg/m2)
Figure 4-16 shows the survival fraction of tumour cells under various administration modes by
applying the pharmacodynamics model described in Equation (3-30). As can be observed, single
infusion is more effective at cell killing in the first 34 hours but less effective than multiple-
infusions afterwards. Comparing different fractionations under the multiple-infusion mode, cell
killing rate is more steady and the fraction of survival tumour cells reaches a lower level as the
0 12 24 36 48 60 72 84 96
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Intr
acel
lula
r C
once
ntr
atio
n i
n T
um
our
(ng/1
05ce
lls)
Time (hr)
single_infusion
twice_infusion
4_times_infusion
8_times_infusion
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
81
number of infusions increases. Results show that the 8-times continuous infusion can improve
the anticancer effectiveness by 7% when compared with a single infusion.
Figure 4-16. Time course of tumour survival fraction under single and multiple continuous infusions.
(dose = 50 mg/m2)
Tumour cell density is determined by the rate of cell killing, cell proliferation and degradation
rates, hence the survival fraction may increase when intracellular concentration falls below a
threshold value during administration intervals. By increasing administration times while
keeping the total administration dose the same, multiple-infusion mode is effective at increasing
doxorubicin concentrations between administration cycles (shown in Figure 4-13, 4-14 and 4-15).
Therefore, the rate of cell death (including cell killing and degradation) is faster than that of cell
proliferation, and as a consequence the fraction of survival tumour cells continuous to fall for a
longer period under multiple-infusion treatments than under a single administration.
Taking results shown in Figures 4-13, 4-14 and 4-15 together, it becomes clear that keeping
doxorubicin at a relatively higher level for a longer period leads to a better anticancer
effectiveness. The treatment outcome is related but not solely determined by the peak extra- and
intra-cellular concentrations. This is because the time-scales of drug transport processes from
blood to the interior of tumour cells are different. Although some of these time scales are
determined by the properties of anticancer drug and tissues, the rate-limiting step is crossing the
cell membrane [55]. The slow cellular influx requires the concentration to sustain at a high level
for a sufficient period of time in order to let more doxorubicin accumulate in the cells.
0 12 24 36 48 60 72 84 96
60
65
70
75
80
85
90
95
100
Surv
ival
Fra
ctio
n (
%)
Time (hr)
single_infusion
twice_infusion
4_times_infusion
8_times_infusion
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
82
Time course profiles of free and bound doxorubicin concentration in the interstitial space of
normal tissue are presented in Figure 4-17. Results indicate that multiple-continuous-infusion
lead to lower peak doxorubicin concentration but a slightly higher concentration level after
administration ends. Compared with doxorubicin intravascular concentration shown in Figure 4-
13, the accumulation of doxorubicin in normal tissue is found to be strongly dependent on
doxorubicin concentration in blood.
0 8 16 24 32 40 48 56 64 72 80 88 96
0.0
2.0x10-5
4.0x10-5
6.0x10-5
8.0x10-5
1.0x10-4
1.2x10-4
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
single_infusion
twice_infusion
4 times_infusion
8 times_infusion
(a)
0 12 24 36 48 60 72 84 96
0.0
5.0x10-5
1.0x10-4
1.5x10-4
2.0x10-4
2.5x10-4
3.0x10-4
3.5x10-4
single_infusion
2 times_infusion
4 times_infusion
8 times_infusion
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
(b)
Figure 4-17. Spatial mean doxorubicin extracellular concentration in normal tissue as a function of time under single
and multiple continuous infusion against time. (a) free doxorubicin, (b) bound doxorubicin (total dose = 50 mg/m2).
Figure 4-18 compares the single infusion with a high dose and multiple-infusions with a low
dose. In Figure 4-18 (a) and (c), intracellular concentration resulted by multiple-infusions with
low dose is obviously low, and quantitative comparisons show that 8 times infusion with half
dose can reduce the peak concentration by an order of magnitude
The survival fractions of tumour cells under different administration modes are shown in Figure
4-18 (b) and (d). Numerical results show that differences in the maximum cytotoxic effect
between high drug dose single infusion and low drug dose 8 times infusion are 8% with a 50%
dose and 3% with a 67% dose for multiple infusions. In addition, the recovery of tumour cells is
slower with low dose 8 times infusion, as shown in Figure 4-18 (d). Consequently, multiple-
continuous-infusion seems to offer improved treatment outcome overall.
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
83
Figure 4-18. Comparison of various doses under single and multiple-infusions. (a) intracellular concentration and (b)
cell survival fraction under single infusion with 50 mg/m2 and multiple infusions with 25 mg/m
2 as a function of
time ; (c) intracellular concentration and (d) cell survival fraction under single infusion with 75 mg/m2 and multiple
infusions with 50 mg/m2 as a function of time.
4.2.3.2. Bolus Injection
The doxorubicin concentration in plasma under various bolus injection schedules are shown in
Figure 4-19 (a). Since the injection duration is short enough to be ignored, peak blood
concentration is reached at the beginning of each administration. Fast plasma clearance is the
reason for the rapid fall of blood concentration after administration. Quantitative comparisons
show that the peak intravascular concentration is inversely related to the number of injections.
This means that multiple bolus injection can reduce the peak concentration in plasma, and hence
may help reduce the risk of cardiotoxicity.
The free and bound doxorubicin concentration in the interstitial space is shown in Figure 4-19 (b)
and (c). Because doxorubicin needs to permeate through the vasculature wall and then enter the
0 12 24 36 48 60 72 84 96
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
50 mg/m2_single_infusion
25 mg/m2_twice_infusion
25 mg/m2_4_times_infusion
25 mg/m2_8_times_infusion
Intr
acel
lula
r C
once
ntr
atin
in T
um
our
(ng/1
05ce
lls)
Time (hr)
(a)
0 12 24 36 48 60 72 84 96
65
70
75
80
85
90
95
100
50 mg/m2_single_infusion
25 mg/m2_twice_infusion
25 mg/m2_4_times_infusion
25 mg/m2_8_times_infusion
Surv
ival
Fra
ctio
n i
n T
um
our
(%)
Time (hr)
(b)
0 8 16 24 32 40 48 56 64 72 80 88 96
0.00
0.25
0.50
0.75
1.00
1.25
1.50
1.75
2.00
2.25
Intr
acel
lula
r C
once
ntr
atio
n i
n T
um
our
(ng/1
05ce
lls)
Time (hr)
75 mg/m2_single_infusion
50 mg/m2_twice_infusion
50 mg/m2_4_times_infusion
50 mg/m2_8_times_infusion
(c)
0 8 16 24 32 40 48 56 64 72 80 88 96
55
60
65
70
75
80
85
90
95
100
Surv
ival
Fra
ctio
n i
n T
um
our
(%)
Time (hr)
75 mg/m2_single_infusion
50 mg/m2_twice_infusion
50 mg/m2_4_times_infusion
50 mg/m2_8_times_infusion
(d)
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
84
interstitial space, its extracellular concentration shares a similar pattern to the intravascular
concentration shown in Figure 4-19 (a).
Figure 4-19. Comparison of doxorubicin concentration under single and multiple bolus injections as a function of
time (total dose = 50 mg/m2). (a) intravascular concentration, (b) free doxorubicin extracellular concentration, (c)
bound doxorubicin extracellular concentration, and (d) intracellular concentration
The predicted time courses of doxorubicin intracellular concentration, for a total dose of 50
mg/m2 administered by different bolus injection schedules, are compared in Figure 4-19 (d).
Results show that single injection leads to the highest peak intracellular concentration, and the
peak value decreases with the number of injections. Although sharing a similar trend to
intravascular concentration, intracellular concentration takes longer to drop to its lowest level
which is approximately zero. This is because the time-scale for crossing cell membrane is much
longer than that of plasma clearance [55].
Figure 4-20 presents the survival fraction of tumour cell based on the predicted intracellular
concentration over time. The cytotoxic effect of single bolus injection on tumour cells is
0 4 8 12 16 20 24 28 32 36 40 44 48
0.000
0.002
0.004
0.006
0.008
0.010
0.012
single_injection
twice_injection
4 times_injection
8 times_injection
Doxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
(kg/m
3)
Time (hr)
(a)
0 6 12 18 24 30 36 42 48
0.0
2.0x10-4
4.0x10-4
6.0x10-4
8.0x10-4
1.0x10-3
1.2x10-3
1.4x10-3
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
single_injection
twice_injection
4 times_injection
8 times_injection
(b)
0 6 12 18 24 30 36 42 48
0.0
1.0x10-3
2.0x10-3
3.0x10-3
4.0x10-3
5.0x10-3
single_injection
twice_injection
4 times_injection
8 times_injection
Bound D
oxourb
icin
Extr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
(c)
0 6 12 18 24 30 36 42 48
0.0
0.5
1.0
1.5
2.0
2.5
single_injection
twice_injection
4 times_injection
8 times_injection
Intr
acel
lula
r C
once
ntr
atio
n i
n T
um
our
(ng/1
05ce
lls)
Time (hr)
(d)
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
85
significantly lower than that of multiple injections, and the difference is up to 15% between the
single and 8-times injections.
Figure 4-20. Tumour survival fraction as a function of time with single and multiple bolus injection
(total dose = 50 mg/m2)
Free and bound doxorubicin concentration in normal tissue under different bolus injection
schedules are compared in Figure 4-21. Results show that peak concentration of doxorubicin in
normal tissue can be reduced by increasing the number of injections.
Figure 4-21. Comparison of doxorubicin concentration under single and multiple bolus injections as a function of
time (total dose = 50 mg/m2). (a) free and (b) bound doxorubicin concentration in normal tissue.
The treatment outcomes under 8-times bolus injection and a single 2-hour continuous infusion
for the total dose of 50 mg/m2 non-encapsulated doxorubicin are compared in Figure 4-22.
Although multiple injections help to improve the anticancer effectiveness as already shown in
0 8 16 24 32 40 48 56 64 72 80 88 96
72
76
80
84
88
92
96
100
single
twice
4 times
8 timesSurv
ival
Fra
ctio
n o
f T
um
our
Cel
ls (
%)
Time (hr)
0 6 12 18 24 30 36 42 48
0.0
1.0x10-4
2.0x10-4
3.0x10-4
4.0x10-4
5.0x10-4
6.0x10-4
single_injection
twice_injection
4 times_injection
8 times_injection
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
(a)
0 6 12 18 24 30 36 42 48
0.0
2.0x10-4
4.0x10-4
6.0x10-4
8.0x10-4
1.0x10-3
1.2x10-3
1.4x10-3
1.6x10-3
single_injection
twice_injection
4 times_injection
8 times_injection
Bound D
oxourb
icin
Extr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
(b)
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
86
Figure 4-20, results obtained with 8-times bolus injection is still not as good as that with 2-hour
continuous infusion for the same dose.
Figure 4-22. Comparison of survival fraction between single continuous infusion and 8 times bolus injection.
(total dose = 50 mg/m2)
4.2.4. Comparison of 2-D and 3-D Models
All the simulations presented so far are based on a 2-D model reconstructed from a selected MR
image of a transverse section. However, a real tumour and its surrounding normal tissues are
three-dimensional (3-D). Hence, it is necessary to evaluate the need for 3D simulation by
quantitative comparison of results based on 2-D and 3-D models.
Figure 4-23. 3-D model geometry.
For this purpose, a 3-D model is reconstructed from the same series of MR images acquired from
a patient with prostate tumour, and a representative MR image is already shown in Figure 4-1 (a).
The reconstructed 3-D geometry is shown in Figure 4-23, where the tumour region is highlighted
0 8 16 24 32 40 48 56 64 72 80 88 96
65
70
75
80
85
90
95
100
Fra
ctio
n o
f S
urv
ival
Tum
our
Cel
ls (
%)
Time (hr)
8-times bolus injection
2-hour continuous infusion
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
87
in orange which is surrounded by normal prostatic tissues in pale blue. The dimension of the
reconstructed 3-D model is approximately 40 mm in length, and the tumour and normal tissue
volumes are 2.28×10-6
m3 and 4.42×10
-5 m
3, respectively. The final mesh consists of 757,453
tetrahedral elements which have been tested to produce grid independent solutions.
Numerical simulation has been carried out for a 2-hour continuous infusion of 50 mg/m2
doxorubicin. Comparisons of drug concentration and drug cytotoxic effect are summarized in
Figure 4-24. The time courses of free doxorubicin extracellular concentration based on the 2-D
and 3-D models are compared in Figure 4-24 (a). No difference can be observed between the two
models. Quantitative differences in drug concentrations and tumour cell density are summarized
in Figure 4-24 (b), which shows that differences between the 2-D and 3-D models are less than
0.25%, justifying the use of 2-D models in this study. These results agree with the work of other
researchers [79]. However, it is important to point out that the high level of consistence between
the 2-D and 3-D models is mainly due to the fact that both drug and tissue properties are
assumed to be uniform in each region. The effect of heterogeneous distribution of
microvasculature in tumour will be addressed in Chapter 5.
Figure 4-24. Comparison of drug concentration and cell density between 2-D and 3-D models. (a) free doxorubicin
extracellular concentration in tumour as a function of time, (b) maximum percentage differences in drug
concentration and tumour cell density (total dose = 50 mg/m2)
4.3. Discussion
After intravenous infusion, anticancer drugs need to experience the following steps before
reaching the interior of tumour cells: clearance from plasma, passing through the
0 1 2 3 4 5 6 7 8 9
0.0
2.0x10-5
4.0x10-5
6.0x10-5
8.0x10-5
1.0x10-4
1.2x10-4
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
2-D Model
3-D Model
(a)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
Dc
Ci_tumour
Cfe_normal_tissue
Cbe_tumour
Cfe_normal_tissue
Per
centa
ge
of
Dif
fere
nce
(%
)
Cfe_tumour
(b)
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
88
microvasculature wall, diffusion and convection in extracellular space and crossing tumour cell
membrane. The time scale for each of these transport steps is different owing to the different
properties of microvasculature, interstitial fluid and the kinetics of cellular uptake. Data in the
literature suggest that there is a delay between intracellular and extracellular drug concentration
levels [55]. This is the reason why extracellular concentration needs to be maintained at a high
level for long enough to enable cellular uptake. Simulation results in this study show that peak
intracellular doxorubicin concentration is achieved 0.8~19 minutes later than the corresponding
peak extracellular concentration. Comparisons between single and multiple administration
methods as well as dose levels suggest that a more effective administration is more successful in
maintaining high levels of extracellular doxorubicin over longer time (shown in Figure 4-5, and
4-10 (b,c)), and thereby raising intracellular levels (shown in Figure 4-8, and 4-10 (d)) and
enhancing cellular uptake.
Anticancer effectiveness is usually evaluated by cell survival fraction, which can be predicted by
different mathematical models. As an improvement of the AUC model [104], the peak
intracellular concentration model [55, 90] was found to have better agreement to experimental
data. However, this model predicts the best cell kill fraction for the entire treatment only, without
giving detailed information on the dynamic variation of tumour cell density over time.
Knowledge of temporal variation of cell survival is important since both extra- and intra-cellular
drug concentrations change with time during a treatment period. The pharmacodynamics model
employed in this study overcomes this shortcoming by providing temporal profiles of cell
survival fraction. In addition, the model allows cell proliferation and degradation to be included,
making it a step closer to mimicking an in vivo environment. With the consideration of cell
proliferation and degradation in the present study, results shown in Figures 4-8, 4-10, 4-14, 4-15
and 4-16 demonstrate that sustained levels of extracellular concentration lead to higher
intracellular concentration over time, which can cause more effective tumour cell killing.
Comparisons of drug concentration and survival fraction of tumour cells between 2-D and 3-D
models show that the difference between the two models is negligible. However, this verdict is
based on a number of assumptions. Firstly, the properties of tumour and normal tissues are
assumed to be constant. These may vary with the tissue type, growth stage and individual
conditions of a patient. Secondly, transport properties and vascular distribution are assumed to be
4. NON-ENCAPSULATED DRUG DELIVERY TO SOLID TUMOUR
89
uniform, which is not true in a real tumour. All these properties will affect the way anticancer
drugs are transported to tumour cells, and eventually the effectiveness of anticancer treatment.
The model parameters adopted in this study correspond to average and representative values for
tumour and normal tissues extracted from the literature, which may be sufficient for qualitative
study, but more information would be required for detailed quantitative analysis.
4.4. Summary
In this chapter, various administration modes, schedules and dose levels are compared based on a
2-D prostate tumour model with a realistic geometry. The results show that better anticancer
effectiveness is achieved with administration methods that are able to maintain a high level of
extracellular concentration over the entire treatment period.
Results on single administrations show that bolus injection is much less effective in tumour cell
killing than continuous infusion, the latter being not only effective in improving the cytotoxic
effect of doxorubicin on tumour cells, but also in reducing peak level of doxorubicin in blood
stream. Modelling results also show that rapid continuous infusion leads to faster cell killing at
the beginning of the treatment, but slow infusion modes perform better over time.
Increasing drug doses can increase concentration levels in all regions of tumour, and improve the
effectiveness of anticancer treatments. Whilst consideration must be given to potential side
effects, which limit the dose used in clinical treatment, the highest tolerable dose gives the best
result.
Multiple-administration can improve the efficacy of anticancer treatments, while significantly
reducing the peak intravascular concentration and therefore lower the risk of cardiotoxicity
caused by doxorubicin in blood stream. Treatment outcomes may be improved by increasing the
number of administrations. The simulation results indicate that reducing the total dose while
increasing the number of continuous infusions is an effective strategy to reduce plasma drug
concentration, hence the associated cardiotoxicity, although at the expense of a slight reduction
in predicted anticancer effectiveness.
5. Effect of Tumour Properties on Drug
Delivery
The properties of tumours are decisive factors in drug transport and final treatment
efficacy. Since in most chemotherapies anticancer drugs are carried via the systemic
circulation, the characteristics of tumour vasculature plays a crucial role in drug
delivery. On the one hand, vasculature in malignant tumours is highly abnormal and
heterogeneous. A typical tumour consists of a necrotic core with no functional blood
vessels in the centre, semi-necrotic regions with less capillaries and veins, and a
tumour advance front which is rich in arterioles, capillaries and venules [49]. On the
other hand, tumour vasculature density varies in the growth process, with less
vasculature in large, advanced tumours.
In this chapter, a liver tumour model incorporating details of microvascular
distribution is adopted first to examine the effect of heterogeneous vasculature
distribution on drug delivery. This is followed by studies of drug transport in prostate
tumours with different sizes and vasculature density.
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
91
5.1. Heterogeneous Vasculature Distribution
The effect of vasculature distribution is investigated in a liver tumour with anticancer drugs
being directly administrated through intravenous infusion. The drug transport processes and their
governing equations are consistent with those used in Chapter 4. The unique tumour geometry
and parameters are described below.
5.1.1. Model Geometry
The geometry of a liver tumour is reconstructed from the post contrast MR images acquired from
a patient using a 3.0-Tesla MR scanner (DISCOVERY MR750, GE, Schenectady, New York,
USA). Multislice anatomical images of the liver were acquired in three orthogonal planes with
GR sequence at three time points after systemic infusion of the tracer Gd-DTPA (gadopentetate
dimeglumine, MW∼0.57 kDa), with each image comprising 256×256 pixels. Other imaging
parameters are given in Table 5-1.
Table 5-1. MR imaging parameters
Parameter
(unit)
Pixel Size
(mm)
Field of View
(mm)
Slice Thickness
(mm)
Repetition Time
(ms)
Echo Time
(ms)
1.3 420 5.0 3.808 1.808
In order to build a 2D model of the tumour and its surrounding normal tissue, a representative
transverse image is chosen, as shown in Figure 5-1 (a). The chosen section covers the maximum
dimension of the tumour. Image processing is carried out by using Mimics (Materialise HQ,
Leuven, Belgium), with the tumour and its sub-regions being segmented based on pixel
intensities and contrast enhancement. It can be seen that the tumour is located at the corner with
a maximum dimension of 41 mm and an area of 9.0×10-4
m2. The tumour is divided into 6 sub-
regions based on differences in signal intensities which are correlated to vessel densities as
described later. Region_1 is further divided into two regions in order to compensate for any
potential boundary effect caused by the lack of surrounding normal tissue. The resulting
smoothed contours of the tumour and normal tissues are imported into ANSYS ICEM CFD to
generate a computational mesh. The dimensions of the reconstructed model are approximately
118.8 mm (maximum width) by 126.8 mm (maximum depth), and the area of the normal tissue is
8.7×10-3
m2. The final mesh consists of 88,594 triangular elements. This is obtained based on
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
92
mesh sensitivity tests which are carried out until the difference in predicted drug concentration
between the adopted mesh and a 10-time finer mesh is less than 1%.
(a)
Normal Tissue
Region_3
Region_5
Region_1_1
Region_4
Region_2
Region_6
Region_1_2
(b)
Figure 5-1. Model geometry: (a) MR image of the liver tumour (in red) and its surrounding tissue (in pule blue); (b)
the reconstructed 2-D geometry.
5.1.2. Model Parameters
A large number of model parameters are required for the computational simulation; these include
physical and transport properties for doxorubicin, tumour and normal tissues. Essential model
parameters adopted in the present study are summarized in Tables 3-1 and 3-3. All the geometric
and transport parameters are considered to be independent of time, since the growth of tumour
and its surrounding normal tissue is negligible during one simulated cycle.
The parameter used to describe the density of vasculature is the surface area of blood vessels per
unit volume of tumour tissue (S/V). Since Gd-DTPA is an approved contrast agent for imaging of
blood vessels and inflamed or diseased tissue with leaky blood vessels, this MRI-visible tracer is
first systemically administrated before MR imaging. The concentration of Gd-DTPA tracer in the
interstitial fluid can be described by
GdpvGdGdpGd CFCC
V
SP
dt
dC__ 1 (5-1)
where t is time, P is the permeability of the tracer, S/V is the surface area of blood vessels per
unit volume of tumour tissue, CGd and Cp represent the tracer concentration in the interstitial
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
93
fluid and blood, respectively, Fv is the filtration rate from the blood vessels per unit volume of
tumour tissue, and σ is the osmotic reflection coefficient of the tracer (σ = 0.6 [136])
The average interstitial concentration of Gd-DTPA has been found to be proportional to the
signal relative intensity enhancement [136]:
1100_
0_
rTS
SSC
Gd
GdGd
Gd
(5-2)
where SGd and SGd_0 represent the signal with and without the tracer, Gd-DTPA. T10 denotes the
relaxation time and r1 is the relaxivity, both of which are constant and can be cancelled out
during the evaluation of S/V.
The spatial-mean signal intensity of each tumour region at three different time points is
calculated, and the relative value of surface area of blood vessel per tumour size, (S/V)*, can be
obtained by interpolation. The ratio of (S/V)* in each tumour region to the average value for the
entire tumour region is then used to calculate values for S/V, by scaling with the standard value
from the literature which is 200 cm-1
. The scaled values are summarized in Table 5-2 in
descending order of microvasculature density.
Table 5-2. The relative and scaled values of surface area of blood vessel per tumour size (cm-1
)
Average Region_1 Region_2 Region_3 Region_4 Region_5 Region_6
(S/V)* 776.52 1741.73 860.84 554.73 502.23 428.69 324.48
S/V 200.00**
448.60 221.72 142.88 129.35 110.41 83.57
** baseline value from literature [49-51]
5.1.3. Numerical Methods
The mathematical equations are solved by means of a finite volume based computational fluid
dynamics (CFD) code, ANSYS FLUENT, together with tailor-made routines using the User
Defined Scalar (UDS) function. The second order UPWIND scheme is adopted for spatial
discretization, while temporal discretization is achieved by employing the second order implicit
backward Euler scheme. Regarding boundary conditions, the external boundaries are the outer
surface of the normal tissue and tumour regions_1_2, where a zero relative pressure is prescribed.
All variables at the internal boundaries between the tumour and normal tissue and between
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
94
different tumour regions are assumed to be continuous. The governing equations for interstitial
fluid flow are solved first, and the resolved pressure and velocity fields in tumour and normal
tissues are then used to solve the mass transfer equations for drug transport. A fixed time-step of
10 seconds is chosen following a time-step sensitivity test. The convergence criteria for residual
tolerances are 1×10-5
for momentum equations and 1×10-8
for mass transfer equations.
5.1.4. Results
A total dose of 50 mg/m2 non-encapsulated doxorubicin [3] is continuously infused within 2
hours [55], simulating a typical treatment for a 70 kg patient [51]. As a foundation for drug
delivery, the micro-mechanical environment in the tumour is examined first. This is followed by
an analysis of the concentration profiles and drug cytotoxic effect.
5.1.4.1. Interstitial Fluid Field
Interstitial fluid pressure (IFP) plays an important role in determining the transport of drug in
tumours. IFP is obtained by solving the governing equations for interstitial flow for the entire
tumour region and its surrounding normal tissue, subject to the boundary conditions described
above and a vascular pressure of 2080 Pa. As shown in Figure 5-2, there is no discernable
difference among different tumour regions, although IFP in the tumour is much higher than that
in the normal tissue. The observation that IFP is not sensitive to tumour microvasculature
distribution is consistent with the results of Baxter and Jain [50] who found that even the
presence of a necrotic core had virtually no effect on the IFP profile, unless the necrotic core size
is greater than 90% of the tumour size. This is because in regions except for a thin layer close to
the outer boundary of the tumour, IFP equilibrates with the effective vascular pressure, which is
given by ivpvp ; refer to Chapter 3 for definition of symbols and parameter values.
Based on the model parameters adopted in this study, the effective vascular pressure is 1533.88
Pa; this is reached in all tumour regions except region_1 and region_2, where the mean IFP is
slightly lower. This means that there is no net filtration between the microvessels and
interstitium in regions 3~6 which are surrounded by regions 1 and 2. On the other hand, the
rather uniform IFP in the interior of tumour implies that pressure-induced convection of
interstitial fluid is weak there, and convective drug transport in the interstitium occurs mainly
within a thin layer at the tumour/normal tissue interface, where there is a steep pressure gradient.
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
95
Figure 5-2. Interstitial fluid pressure distribution in tumour and normal tissues.
5.1.4.2. Doxorubicin Concentration
Figure 5-3 shows the spatial distribution of free doxorubicin at different time points in the
tumour. It is clear that the extracellular concentration of doxorubicin is non-uniform, which is
not surprising since the transport of drug is channelled through blood vessels, and as such,
vasculature density is expected to have a strong influence on the spatial distribution of drug.
Since the difference in extra- and intra-cellular concentrations between sub-regions 1_1 and 1_2
is very small, these sub-regions are combined and referred to as region_1 in the following
discussion.
1 hour 3 hour 6 hour 12 hour
Figure 5-3. Spatial distribution of free doxorubicin extracellular concentration in tumour regions (total dose = 50
mg/m2, infusion duration = 2 hours)
Figure 5-4 shows the variation of extracellular concentration of free and bound doxorubicin in
the liver tumour and each of its sub-regions of different vessel densities. In all regions, the
predicted concentration increases during the infusion period, when doxorubicin is administrated
continuously into the circulatory system. After the end of administration, both free and bound
1500.0
1236.2
1018.8
839.6
691.9
570.2
470.0
387.2
319.2
263.0
216.8
178.7
147.2
121.3
100.0
IFP (Pa)
8.25×10-5
6.63×10-5
5.33×10-5
4.28×10-5
3.44×10-5
2.76×10-5
2.22×10-5
1.79×10-5
1.44×10-5
1.15×10-5
9.27×10-6
7.45×10-6
Cfe (kg/m3)
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
96
doxorubicin concentrations start falling immediately. There are obvious differences in the drug
concentrations in the different tumour regions until at approximately 12 hours, when uniform
doxorubicin concentrations are reached in all tumour regions (also demonstrated in Figure 5-3).
The rate of change of doxorubicin concentration is slower in regions with low vessel density (e.g.
region_6), owing to reduced drug exchange between plasma and interstitial fluid as a result of
sparse vasculature.
0 1 2 3 4 5 6 7 8 9 10 11 12
0.0
2.0x10-5
4.0x10-5
6.0x10-5
8.0x10-5
1.0x10-4
1.2x10-4
Region_1
Region_2
Region_3
Region_4
Region_5
Region_6
Fre
e D
oxoru
bic
in E
xtr
acel
lual
r C
once
ntr
atio
n
(kg/m
3)
Time (hr)
(a)
0 1 2 3 4 5 6 7 8 9 10 11 12
0.0
5.0x10-5
1.0x10-4
1.5x10-4
2.0x10-4
2.5x10-4
3.0x10-4
3.5x10-4
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
(kg/m
3)
Time (hr)
Region_1
Region_2
Region_3
Region_4
Region_5
Region_6
(b)
Figure 5-4. Spatial mean doxorubicin extracellular concentration in tumour as a function of time. (a) Free
doxorubicin, and (b) bound doxorubicin. (total dose = 50 mg/m2, infusion duration = 2 hours)
Although vascular densities vary in the different tumour regions, the results in each tumour
region show that nearly 75% doxorubicin is in bound form, and both free and bound drugs share
a similar time course, confirming that the mechanism of binding with protein is independent of
vasculature distribution.
0 1 2 3 4 5 6 7 8 9 10 11 12
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Region_1
Region_2
Region_3
Region_4
Region_5
Region_6
Doxoru
bic
in I
ntr
acel
lula
r C
once
ntr
atio
n
(ng/1
05ce
lls)
Time (hr)
Figure 5-5. Spatial mean of doxorubicin intracellular concentration in tumour regions as a function of time. (total
dose = 50 mg/m2, infusion duration = 2 hours)
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
97
Intracellular concentration also displays a non-uniform pattern among the tumour regions of
different vessel densities. As shown in Figure 5-5, the rate of change of intracellular
concentration is slower and peak concentration is lower in poorly vascularised regions, similar to
the pattern observed for extracellular concentration. This is because cellular uptake follows the
process of transvascular transport and migration in interstitial space, therefore, the intracellular
concentration profile is strongly dependent on the concentration in extracellular space. The
highest peak is observed in the region with the densest microvasculature, and results in Figure 5-
5 indicate that the peak value increases monotonically with vascular density.
5.1.4.3. Doxorubicin Cytotoxic Effect
The predicted anticancer effectiveness, assessed based on the variation in the percentage of
tumour cells (Figure 5-6), suggests that cell killing is more effective in well-vascularised regions
in the first few hours, but the trend is reversed thereafter. However, the difference among the
various regions is small, with the maximum difference being 3% between region_1 (high vessel
density) and region_6 (low vessel density) at 12-hour following drug administration.
0 2 4 6 8 10 12
75
80
85
90
95
100
Region_1
Region_2
Region_3
Region_4
Region_5
Region_6
Surv
ival
Fra
ctio
n o
f T
um
our
Cel
ls (
%)
Time (hr)
Figure 5-6. Spatial mean percentage of tumour cells in tumour regions as a function of time (total dose = 50 mg/m2,
infusion duration = 2 hours).
The spatial distributions of survival cell fraction in the tumour regions at different time points are
shown in Figure 5-7. As a result of the non-uniform distribution of extracellular concentration
(see Figures 5-3 and 5-4), different treatment outcomes are found in different tumour regions.
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
98
98.7
95.4
92.0
88.6
85.2
81.8
78.4
75.0
f (%)
1 hour 3 hours 6 hours 12 hours
Figure 5-7. Spatial distribution of survival cell fraction in tumour regions.
(total dose = 50 mg/m2, infusion duration = 2 hours)
5.1.4.4. Comparison of Various Administration Schedules
Comparisons between different doses (25, 50 and 75 mg/m2) for a 2-hour continuous infusion
confirm that the qualitative trend presented above is not altered by changes in dose level. As
shown in Figure 5-8, the maximum differences in free doxorubicin extra- and intra-cellular
concentrations among tumour regions increase proportionally with an increase in drug dose.
0.0
1.0x10-5
2.0x10-5
3.0x10-5
4.0x10-5
5.0x10-5
6.0x10-5
75 mg/m2
50 mg/m2
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
(kg/m
3)
25 mg/m2
(a)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Intr
acel
lula
r C
once
ntr
atio
n (
ng/1
05ce
lls)
25 mg/m2
50 mg/m2
75 mg/m2
(b)
Figure 5-8. Maximum difference among tumour regions over treatment with various doses. (a) free doxorubicin
extracellular concentration and (b) intracellular concentration (infusion duration = 2 hours)
Comparisons are also made between different infusion durations (1-hour, 2-hour and 4-hour) for
a 50 mg/m2 continuous infusion, and the qualitative tread among tumour regions is not altered by
this parameter. The maximum differences in free doxorubicin extra- and intra-cellular
concentration among tumour regions, shown in Figure 5-9, indicate that prolonging the infusion
duration may help to reduce the difference in both extra- and intra-cellular concentration among
tumour regions.
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
99
0.0
2.0x10-5
4.0x10-5
6.0x10-5
8.0x10-5
1.0x10-4
1.2x10-4
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
(kg/m
3)
1 hour 2 hour 4 hour
(a)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Intr
acel
lula
r C
once
ntr
atio
n (
ng/1
05ce
lls)
1 hour 2 hour 4 hour
(b)
Figure 5-9. Maximum difference among tumour regions over treatment with various infusion durations. (a) free
doxorubicin extracellular concentration and (b) intracellular concentration (total dose = 50 mg/m2)
5.2. Tumour Size
The effect of tumour size is investigated using three prostate tumours under free form drug
delivery. The drug transport processes and their governing equations are consistent with those
adopted in Chapter 4. The unique tumour geometry and parameters are described below.
5.2.1. Model Geometry
Three prostate tumour models are reconstructed from images acquired from patients using a 3.0-
Tesla MR scanner (DISCOVERY MR750, GE, Schenectady, New York, USA). Multislice
anatomical images of the prostate were acquired in three orthogonal planes with echo-planer (EP)
sequence, with each image comprising 256 by 256 pixels. Other imaging parameters are given in
Table 5-3.
Table 5-3. MR imaging parameters
Pixel Size
(mm)
Field of View
(cm)
Slice Thickness
(mm)
Repetition Time
(ms)
Echo Time
(ms)
case_1
1.250 32.0 7.000
3675 85.7
case_2 4000 84.4
case_3 4000 91.5
An example of the MR images has been shown previously in Figure 4-1. Transverse images are
processed using image analysis software Mimics (Materialise HQ, Leuven, Belgium), and the
tumours are segmented from its surrounding normal tissues based on signal intensity values. The
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
100
resulting smoothed surfaces of the tumour and normal tissues are imported into ANSYS ICEM
CFD to generate computational mesh for the entire volume. The dimension of the reconstructed
models shown in Figure 5-10 is approximately 40~50 mm in length for three cases. The final
mesh consists of 1,173,908, 1,163,374 and 1,474,027 tetrahedral elements for case 1, 2 and 3,
respectively. These meshes have been tested to produce grid independent solutions.
tumour normal tissue
Figure 5-10. Model geometry: (a) case_1; (b) case_2; (c) case_3
5.2.2. Model Parameters
Since the time window of simulations is smaller than the growth of tumour, all the geometric and
transport parameters are considered to be independent of time. Baseline values for all model
parameters are summarized in Tables 3-1 and 3-3. The tumour size dependent parameter is the
surface area of blood vessels per unit volume of tumour tissue (S/V), which reflects the
microvasculature density.
Figure 5-11. Estimation of blood vessel surface area to tissue volume ratio of tumours with various tumour sizes
(experimental data extracted from [116]). For the best fitted curve using equation y=axb (a = 54.68, b = -0.2021).
0 200 400 600 800 1000 1200 1400 1600
10
12
14
16
18
20
22
24
26
28
S/V
(m
m-1)
Tumour Volume (mm3)
experimental data
fitting curve
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
101
Pappenheimer measured S/V normal tissue and found this to be 70 cm-1
[117]; this value was
adopted by Baxter and Jain [49, 50, 88] in their series of works. Hilmas and Gillette [116]
measured this ratio in tumours with different volumes and at different growth stages by using
morphometric methods. By using non-linear least squares curve-fitting technique, the
relationship between S/V and tumour size is obtained (shown in Figure 5-11), and the S/V value
for each prostate tumour is then calculated based on its own size; all the values obtained are
summarized in Table 5-4.
Table 5-4. Tumour size and the blood vessel surface area to tissue volume ratio in each model
Tumour size (mm3) S/V (mm
-1)
case_1 45.46 25.28
case_2 2277.97 11.46
case_3 7118.84 9.10
The numerical method and boundary conditions are consistent with previous studies in this
Chapter.
5.2.3. Results
A numerical simulation has been carried out for continuous infusion of 50 mg/m2 [3] non-
encapsulated doxorubicin in 2 hours [55], which corresponds to a standard treatment for a patient
of 70 kg body weight [51]. The obtained interstitial fluid field is presented first as this provides
the microenvironment for drug delivery. This is followed by comparisons of doxorubicin
concentration in the three prostate tumours to investigate the effect of tumour size on drug
delivery. Finally, predicted treatment outcomes based on survival fractions of tumour cells
calculated from the pharmacodynamics model are compared.
5.2.3.1. Interstitial Fluid Field
Because of abnormalities in tumour, such as vasculature, cellular matrix and the components of
interstitial fluid, tumour interstitial fluid pressure (IFP) has been found to be higher than that in
normal tissues [49, 50, 88, 89]. Predicted IFP contours at a representative cross-section for each
of the three tumours are shown in Figure 5-12. IFP is uniformly high in the entire tumour except
in a thin layer close to the tumour boundary. Owing to the low pressure in normal tissue, a sharp
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
102
pressure gradient can be found near the boundary between tumour and the surrounding normal
tissue.
148013201160100084068052036020040
IFP (Pa)
case_1 case_2 case_3
Figure 5-12. Interstitial fluid pressure distribution in tumour and normal tissue
The predicted spatial mean IFP and spatial mean transvascular flux per tumour volume in each
tumour are shown in Figure 5-13 (a) and (b), respectively. It can be clearly observed that spatial
mean IFP varies with tumour size, with higher IFP in larger tumour. Since IFP in tumour centre
are at the same for all three models (at 1533.88 Pa), the observed variation of mean IFP with
tumour size is caused by the thin layer with a sharp IFP gradient near the tumour boundary as
shown in Figure 5-12. It has been reported in a parameter sensitivity study that IFP gradient is
steeper in larger tumour [49], which would result in a higher spatial mean IFP. This is consistent
with the finding presented here.
0
100
200
300
1300
1400
1500
1600
45.46 mm3
1478.04 Pa1471.65 Pa
7118.74 mm3
2277.97 mm3
Spat
ical
-mea
n I
nte
rsti
tial
Flu
id P
ress
ure
(P
a)
1399.99 Pa
(a)
0.0
1.0x10-5
2.0x10-5
3.0x10-5
4.0x10-5
5.0x10-5
6.0x10-5
7.0x10-5
8.0x10-5
1.07×10-5 s
-11.50×10
-5 s
-1
Tra
nsv
ascu
lar
Flu
x p
er T
um
our
Volu
me
(s-1)
7.11×10-5 s
-1 (b)
45.46 mm3
2277.97 mm3
7118.74 mm3
Figure 5-13. The spatial mean interstitial fluid pressure and transvacular flux per tumour volume as a function of
tumour size
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
103
The transvascular flux per tumour volume from blood to interstitium (Fv) is governed by
Equation (3-2), which is recalled here:
ivivvv ppV
SKF (3-2)
As shown in Figure 5-13 (b), the spatial mean transvascular flux per tumour volume from
vasculature is higher in small tumour. This is because: (1) the driving force is high in small
tumour owing to the low value of spatial mean interstitial fluid pressure and uniform
intravascular pressure; (2) higher S/V in small tumour improves this exchange. However,
referring to the spatial distribution of IFP in Figure 5-12, the transvascular flux mainly occurs in
the thin layer at tumour/normal tissue interface owing to the equilibrium between IFP and
effective vascular pressure within tumour centre. Results also show that spatial mean IFP and
transvascular flux per tumour volume are non-linearly related to tumour volume.
5.2.3.2. Doxorubicin Concentration
Intravascular concentrations of free and bound doxorubicin are shown in Figure 5-14.
Doxorubicin in blood circulation system keeps increasing in the 2-hour infusion period and
reaches the peak at the end of administration. This is followed by a rapid fall, after which there is
a gradual and slow reduction of doxorubicin concentration as time proceeds. Quantitative
comparison shows that 75% doxorubicin is bound with proteins in blood stream.
0 1 2 3 4 5 6 7 8 9
0.0
1.0x10-4
2.0x10-4
3.0x10-4
4.0x10-4
5.0x10-4
6.0x10-4
7.0x10-4
Doxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n (
kg/m
3)
Time (hr)
free doxorubicin
bound doxorubicin
Figure 5-14. Free and bound doxorubicin concentration in blood after administration as a function of time. (infusion
duration=2 hours, total dose=50mg/m2)
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
104
Free and bound doxorubicin extracellular concentration in tumour and its holding normal tissue
are shown in Figure 5-15. Regardless of the tumour size, both free and bound doxorubicin
concentrations in tumour and normal tissue increase rapidly during the initial period after
administration, reach their peaks and then fall to a low level after the end of infusion. The rate of
change in doxorubicin concentration slows down with the increase in tumour size. This is
because S/V is lower in larger tumour, leading to less drug exchange between blood and tumour
interstitium. As shown in Figure 5-15 (c) and (d), variation of tumour size has no obvious
influence on drug concentration in normal tissue.
0 1 2 3 4 5 6 7 8 9
0.0
2.0x10-5
4.0x10-5
6.0x10-5
8.0x10-5
1.0x10-4
1.2x10-4
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
case_1
case_2
case_3
(a)
0 1 2 3 4 5 6 7 8 9
0.0
5.0x10-5
1.0x10-4
1.5x10-4
2.0x10-4
2.5x10-4
3.0x10-4
3.5x10-4
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
case_1
case_2
case_3
(b)
0 1 2 3 4 5 6 7 8 9
0.0
2.0x10-5
4.0x10-5
6.0x10-5
8.0x10-5
1.0x10-4
1.2x10-4
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
case_1
case_2
case_3
(c)
0 1 2 3 4 5 6 7 8 9
0.0
5.0x10-5
1.0x10-4
1.5x10-4
2.0x10-4
2.5x10-4
3.0x10-4
3.5x10-4
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
case_1
case_2
case_3
(d)
Figure 5-15. Spatial mean doxorubicin extracellular concentration as a function of time under 2-hour continuous
infusion, dose=50mg/m2. (a) Free and (b) bound doxorubicin in tumour, (c) bound and (d) bound doxorubicin in
normal tissue
Figure 5-16 shows the amount of drug exchange between vasculature and interstitial fluid by
convection in each tumour. Compared with the blood concentration in Figure 5-14, doxorubicin
exchange by convection follows the same pattern. This means that blood concentration has a
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
105
direct influence on transvascular transport of doxorubicin by convection. Result in Figure 5-16
also shows that transvascular transport in small tumour is higher than in large tumour.
0 1 2 3 4 5 6 7 8 9
0.0
2.0x10-9
4.0x10-9
6.0x10-9
8.0x10-9
Fre
e D
oxoru
bic
in E
xch
ange
bet
wee
n B
lood
and I
nte
rsti
tial
Flu
id (
kg/m
3)
Time (hr)
case_1
case_2
case_3
(a)
0 1 2 3 4 5 6 7 8 9
0.0
2.0x10-9
4.0x10-9
6.0x10-9
8.0x10-9
1.0x10-8
1.2x10-8
1.4x10-8
1.6x10-8
Bound D
oxoru
bic
in E
xch
ange
bet
wee
n B
lood
and I
nte
rsti
tial
Flu
id (
kg/m
3)
Time (hr)
case_1
case_2
case_3
(b)
Figure 5-16. Doxorubicin exchange between blood vasculature and interstitial fluid by convection as a function of
time. (a) free and (b) doxorubicin exchange (2-hour infusion, total dose =50 mg/m2)
Regardless of the tumour size, transvascular transport of doxorubicin by diffusion shown in
Figure 5-17 experiences the following stages: (1) increasing rapidly until reaching the peak, and
then decreasing slowly to a positive level at the end of administration; (2) dropping sharply to a
negative level in a few minutes and then returning very slowly to zero.
0 1 2 3 4 5 6 7 8 9
-3.0x10-6
-2.0x10-6
-1.0x10-6
0.0
1.0x10-6
2.0x10-6
3.0x10-6
4.0x10-6
5.0x10-6
case_1
case_2
case_3
Fre
e D
oxoru
bic
in E
xch
ange
bet
wee
n B
lood
and I
nte
rsti
tial
Flu
id (
kg/m
3)
Time (hr)
(a)
0 1 2 3 4 5 6 7 8 9
-3.0x10-8
-2.0x10-8
-1.0x10-8
0.0
1.0x10-8
2.0x10-8
3.0x10-8
4.0x10-8
Bound D
oxoru
bic
in E
xch
ange
bet
wee
n B
lood
and I
nte
rsti
tial
Flu
id (
kg/m
3)
Time (hr)
case_1
case_2
case_3
(b)
Figure 5-17. Doxorubicin exchange between blood vasculature and interstitial fluid as a function of time in each
tumour. (a) free and (b) bound doxorubicin exchange by diffusion owing to concentration gradient (2-hour infusion,
total dose = 50 mg/m2)
Because the initial value of doxorubicin extracellular concentration is zero, the rapid increase in
intravascular concentration results in a large concentration gradient between blood and the
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
106
interstitial fluid after administration begins, therefore, transvascular transport by diffusion
increases rapidly to its peak. Since the increase of blood concentration slows down, the gradient
becomes small and leads to a reduction in diffusive transport. However, it is still positive until
the end of drug administration, after which blood concentration drops rapidly and no more
doxorubicin is supplied. As a consequence, the concentration gradient reduces to zero quickly.
Since doxorubicin is washed out quickly, its extracellular concentration becomes higher than
blood concentration, causing the exchange to reverse owing to a negative concentration gradient.
Afterwards, the negative gradient becomes smaller as the extracellular concentration decreases,
causing the diffusive transport to return to zero gradually.
As shown in Figures 5-14 and 5-15, most of doxorubicin is in bound form in blood plasma and
interstitial fluid. However, comparison of trans-vascular exchange of free and bound doxorubicin
presented in Figures 5-16 and 5-17 indicates that transvascular exchange mainly depends on free
doxorubicin. This can be attributed to the small osmotic reflection coefficient and high
permeability of free doxorubicin, which ease the transvascular process. Results also show that
the rate of change slows down as the tumour size increases. This is because the sparse
microvasculature in large tumour reduces drug exchange between blood and interstitial fluid in
tumour.
Comparisons are made between spatial mean convection exchange and diffusion exchange by
applying the same administration method to case_2. The result in Figure 5-18 shows that
diffusion exchange is up to 1000 fold higher than convection during the infusion period. This
means that diffusion dominants the drug transport process in the initial hours of anticancer
treatment. After the end of administration, diffusive transport driven by concentration difference
weakens but is still dominant, resulting in continued reduction in extracellular concentration. It is
worth noting that transvascular transport by convection mainly occurs in a thin layer at the
interface between tumour and normal tissue. On the contrary, transvascular transport by diffusion
takes place in the entire tumour and normal tissue owing to the concentration gradient across
microvasculature walls.
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
107
0 1 2 3 4 5 6 7 8 9
-2.0x10-6
-1.0x10-6
0.0
1.0x10-6
2.0x10-6
3.0x10-6
Fre
e D
oxoru
bic
in E
xch
ange
bet
wee
n B
lood
and I
nte
rsti
tial
Flu
id (
kg/m
3)
Time (hr)
by convection
by diffusion
(a)
0 1 2 3 4 5 6 7 8 9
-1.5x10-8
-1.0x10-8
-5.0x10-9
0.0
5.0x10-9
1.0x10-8
1.5x10-8
2.0x10-8
by convection
by diffusion
Bound D
oxoru
bic
in E
xch
ange
bet
wee
n B
lood
and I
nte
rsti
tial
Flu
id (
kg/m
3)
Time (hr)
(b)
Figure 5-18. Comparison of (a) free and (b) bound doxorubicin exchange between blood and interstitial fluid by
convection and diffusion in case_2 (2-hour infusion, total dose = 50mg/m2)
Figure 5-19 presents the intracellular doxorubicin concentration in the tree tumours for 2-hour
direct infusion. Given that intracellular concentration depends on the free doxorubicin
extracellular concentration in tumour, intracellular concentration increases rapidly to its peak and
before falling off. The computational model predicts that a small tumour with denser
microvasculature has more rapid changing rates and higher peaks. However, intracellular
concentration in a large tumour sustains at a slightly higher level over time.
0 1 2 3 4 5 6 7 8 9
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Doxoru
bic
in I
ntr
acel
lula
r C
once
ntr
atio
n i
n T
um
our
(ng/1
05ce
lls)
Time (hr)
case_1
case_2
case_3
Figure 5-19. Temporal profiles of predicated intracellular concentration in tumours with different sizes
(2-hour infusion, total dose = 50 mg/m2)
5.2.3.3. Doxorubicin Cytotoxic Effect
Figure 5-20 presents the fraction of survival tumour cells by applying the pharmacodynamics
model described by Equation (3-30). In small tumours with a denser vasculature, more tumour
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
108
cells can be killed in the first few hours after administration begins. However, the drug cytotoxic
effect in small tumours becomes worse as time proceeds. This is because the intracellular
concentration in small tumours increases rapidly to a higher peak value, but also decreases faster
to a low level (shown in Figure 5-19). This fast reduction leads to insufficient cell killing in the
later hours of a treatment in small tumours. As can be observed, differences in cytotoxic
effectiveness are limited, with a difference of up to 3.2% between case_1 (small tumour) and
case_3 (large tumour) in the simulation duration.
0 1 2 3 4 5 6 7 8 9
75
80
85
90
95
100
Surv
ival
Fra
ctio
n o
f T
um
our
Cel
ls (
%)
Time (hr)
case_1
case_2
case_3
Figure 5-20. Survival cell fraction of tumour as a function of time in three tumours with different sizes (2-hour
infusion, total dose = 50 mg/m2)
5.3. Discussion
Results from the mathematical model show that IFP is uniform within a tumour except for the
tumour boundary because IFP in the tumour interior reaches equilibrium with the effective
vascular pressure, implying that there is no net filtration between the blood vessels and tumour
interstitium. Therefore, convective transvacular transport of drug only occurs within a thin layer
at the tumour/normal tissue interface, where there is a steep pressure gradient, and hence
transvascular transport mainly replies on diffusion driven by concentration gradient across
microvasculature wall.
After administration, doxorubicin accumulates faster in tumour regions with denser vasculature
to achieve a higher peak. This is because the rate of transvascular transport is higher than that in
tumour regions with sparser microvasculature. Similarly, doxorubicin is rapidly drained back to
blood lumen after the end of administrations in these regions owing to the fast plasma clearance
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
109
and reverse concentration gradient, resulting in a quick reduction of concentration. Similar
results have been found when examining the effect of tumour size with different
microvasculature density. The rate of change of concentration is faster in small tumours with
dense microvasculature, where the anticancer efficacy is reduced.
Results demonstrate that vasculature in tumour is a crucial factor in drug delivery. Nevertheless,
the mathematical model employed here involves a number of assumptions. First, the tumour
vascular network is treated as a distributed source term in the governing equations instead of
being modelled explicitly. As a result, the effects of intravascular transport and geometric
features of tumour vasculature are not included. While this may be an over-simplification, it is
practical given the lack of information on realistic tumour vasculature geometry and the
complexity in direct modelling of blood flow in a vascular network coupled with interstitial drug
transport. Second, all the tumour regions are assumed to be perfused via microvasculature in the
studied liver tumour model. Drug transport in tumour regions without functional blood vessels,
such as necrotic cores, is not investigated. Third, a 2D tumour model is used instead of a full 3D
model. Based on the comparisons between 2D and 3D models of a prostate tumour discussed in
Chapter 4, the 3D effect on spatial-mean parameters, such as intracellular drug concentration and
tumour cell density, is negligible. This is consistent with the work of Teo et al [79] who found
qualitatively similar results between their 2D and 3D models of a brain tumour.
Other assumptions include: (1) signal intensities of tracer Gd-DTPA in post-contrast MR images
are used to calculate vascular density, with the tracer concentration in the plasma being assumed
to be proportional to the relative change in signal intensity. This could be improved by
measuring the intravascular concentration of Gd-DTPA directly in MR imaging [136]. (2)
uniform transport properties and uniform tumour cell density are assumed in each tumour region.
(3) the intratumoural formation of new blood vessels is a dynamic and complex process
involving several cellular and subcellular events in vivo [137], and the administrated doxorubicin
can inhibit angiogenesis rather than cytotoxic effect on tumour cells in doxorubicin-insensitive
tumour models [138-140]. Changes in blood vessels are expected to directly affect drug
concentration in interstitial fluid and cell killing. However, since these variations usually take
place in days [137, 138] which are much longer than the simulated time window, the
5. EFFECT OF TUMOUR PROPERTIES ON DRUG DELIVERY
110
vasculature-related parameters are assumed to be constant in this series of work, but their time-
varying nature needs to be considered when long simulations are carried out.
5.4. Summary
The effects of vasculature-related microenvironment in tumour have been investigated.
Modelling of a liver tumour with heterogeneous vasculature distribution shows that the tumour
interstitial fluid pressure is insensitive to the distribution of blood vessels. Even in parts of the
tumour that is not well surrounded by holding normal tissue, the pressure gradient there is still
limited to a thin layer with a negligible impact on the tumour interior. Better anticancer
effectiveness is achieved in regions with low vasculature density, owing to the reduced loss of
drugs to the plasma.
Modelling of three prostate tumours with different sizes under direct continuous infusion
demonstrates the nonlinear relationships between spatial-mean interstitial fluid pressure and
tumour volume, as well as between transvascular flux per tumour volume and tumour volume.
During the infusion period, transvascular drug exchange mainly depends on diffusion, owing to
the concentration gradient of free drug between blood and interstitial fluid.
6. Thermo-Sensitive Liposome-Mediated
Drug Delivery to Solid Tumour
Thermo-sensitive liposome-mediated drug delivery is a promising method to reduce
the risk of side effects in anticancer treatment. For well-designed thermo-sensitive
liposomes (TSL) that are stable in blood, encapsulated drugs cannot be released until
local temperature is raised above the phase transition temperature of the liposome
membrane. Among several heating methods that are currently available, high
intensity focused ultrasound (HIFU) is favoured for its non-invasive nature and
accuracy when used under the guidance of MRI. In the focus region of HIFU, both
tumour and blood are heated directly by the absorbed acoustic power, whereas
temperature elevation in the surrounding region is achieved via heat transfer through
convection and conduction.
In this chapter, the acoustic model for HIFU described in Chapter 3 is firstly
validated by comparison with experimental data. Subsequently, the coupled bioheat
transfer and drug transport models are used to investigate TSL-mediated drug
delivery to a realistic 2-D prostate tumour. Finally, a comparison of temperature
profiles between 2-D and 3-D tumour models is presented.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
112
6.1. Acoustic Model for High Intensity Focused Ultrasound
Given that temperature elevation and the following drug release are consequences of absorption
of ultrasound energy in tissue and blood, the acoustic pressure generated by HIFU is an essential
input to the simulation of TSL-mediated drug delivery. The acoustic HIFU model described in
Chapter 3 is applied to a standard test case for which experimental measurements are available.
1.17
1.01
0.85
0.69
0.54
0.39
0.24
0.08
Acoustic Pressure
(MPa)
Figure 6-1. Spatial distribution of acoustic pressure
The test case is for a single element HIFU transducer with a 120-mm focal length and 120-mm
aperture. The transducer with a fixed frequency of 1.33 MHz is driven by a continuous wave to
generate 1.45 MPa at the focus point, with a corresponding intensity of 141.78 W/cm2. Moving
the transducer along the beam and its orthogonal directions, the acoustic pressure in water was
measured by a chosen 0.4 mm hydrophone in Solovchuk’s experiment [114]. The validation
study of acoustic pressure is based on a square geometry (8cm×5.5cm) with the focus region
located in the centre. The spatial distribution of acoustic pressure is shown in Figure 6-1.
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
Numerical Result
Experimental Result
Norm
aliz
ed A
coust
ic P
ress
ure
Axial Distance (mm)
(a)
-5 -4 -3 -2 -1 0 1 2 3 4 5
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
Numerial Result
Experimental Result
Norm
aliz
ed A
coust
ic P
ress
ure
Radial Distance (mm)
(b)
Figure 6-2. Measured and computed acoustic pressure profiles at 1.33 MHz and 1.45 MPa pressure at the focal point.
(a) along axis and (b) radial distance. Experimental data are extracted from [114].
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
113
Normalized by the pressure at the focal point, acoustic pressure along the axial and radial
distance in the focus plane is shown in Figure 6-2. Results show that the predicted acoustic
profiles agree well with the experimental data, demonstrating the validity of the HIFU model for
prediction of acoustic profiles in biological media [114].
6.2. Thermo-Sensitive Liposome-Mediated Delivery
The combined HIFU, bioheat transfer and drug transport models are applied to a prostate tumour,
which is heated simultaneously with the administration of TSL encapsulated doxorubicin. By
solving the governing equations described in Chapter 3, spatial and temporal distributions of
temperature and drug concentration in blood, tumour and normal tissues are obtained. The
expected therapeutic efficacy is evaluated by tumour cell survival fraction.
6.2.1. Model Description
The mathematical model for TSL-mediated drug delivery consists of two parts: bioheat transfer
model and drug transport model. The former covers acoustic energy generated by a single
element ultrasound transducer, heat absorption by blood, tumour and normal tissues, as well as
heat transfer between these compartments; while the later describes the transport of encapsulated,
free and bound doxorubicin in blood, extracellular space of tumour and surrounding normal
tissues, as well as cellular uptake of the drug.
6.2.1.1. Model Geometry
A 2-D model of prostate tumour is reconstructed from MR images; the same model has been
used in Chapter 4. As shown in Figure 6-3, two focus regions are chosen in order to achieve
adequate coverage of the tumour [76, 77].
Figure 6-3. Model geometry and locations of focus regions.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
114
ANSYS ICEM CFD is used to generate the computational mesh. The final mesh consists of
64,966 triangular elements. This is obtained based on mesh independence tests which show that
the difference in predicted drug concentration between the adopted mesh and a 10 times finer
mesh is less than 1%.
6.2.1.2. Model Parameters
Parameters describing the basic physical and transport properties of doxorubicin, tumour and
normal tissue have been given in Tables 3-1, 3-2 and 3-3 in the preceding chapter. Additional
parameters required for the acoustic and bioheat transfer models are summarised in Table 6-1.
Since temperature elevations in response to HIFU heating may influence some of the properties
used in the drug transport model, temperature dependence of these properties is also considered.
Table 6-1. Acoustic and thermal properties
Symbol Description Unit Value
Source Blood Tumour Normal Tissue
va ultrasound speed m/s 1540 1550 1550 [74]
ρ density kg/m3 1060 1000 1055 [74]
c specific heat J/kg/K 3770 3800 3600 [74]
k thermal conductivity W/m/K 0.53 0.552 0.512 [74]
α absorption coefficient 1.5 f 9 f 9 f [74]
wb0 perfusion rate of blood flow at 37oC s-1 - 0.002 0.018 [45]
R universal gas constant J/mol/K - 8.314 8.314 [141]
ΔE activation energy J/mol - 6.67×105 6.67×105 [141]
Af frequency factor s-1 - 1.98×10106 1.98×10106 [141]
* f represents ultrasound frequency
HIFU Transducer
The acoustic source is a single-element, spherically focused transducer, and its parameters are
summarized in Table 6-2. The local acoustic intensity and peak negative pressure are set as 120
W/cm2 and 1.98 MPa, respectively.
Table 6-2. HIFU transducer parameters
Parameter
(unit)
Inside diameter
(mm)
Outside diameter
(mm)
Focal length
(mm)
Frequency
(MHz)
20.0 70.0 62.64 1.10
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
115
Temperature controller
In TSL-mediated drug delivery, it is desirable to keep the tumour temperature within a narrow
hyperthermia range which should not exceed 43oC to avoid vascular shut down [142]. The phase
transition temperature of liposome strongly depends on its formulation, and a typical temperature
is 42oC based on experimental studies [69, 143]. In order to achieve the target temperature range,
a feedback system with an appropriate temperature control mechanism is required. In the present
study, a temperature controller is incorporated, which is designed to adjust the applied ultrasound
power according to the maximum temperature in the heated region (Tmax). The transducer power
is determined by
tTTktTTktP ettciettcp maxargmaxarg (6-1)
where kcp and kci are Proportional-Integral (PI) control parameters, which are set as 5 and 0.006
respectively, based on values in the literature [72] and previous studies.
Permeability
(1) Free & bound doxorubicin
Dalmark and Strom [144] measured the permeability of free doxorubicin at various temperatures,
and found that the logarithmic value of permeability is proportional to temperature. Based on
their data, a linear relationship was found between fold increase in logarithmic value of
permeability and temperature, as given by Equation (6-2), and the best fitting line is shown in
Figure 6-4.
21log zTzP (6-2)
where z1 and z2 are fitted parameters which are 0.07 and 2.43, respectively.
Hence the fold increase in permeability is
211102
1 TTz
PP
(6-3)
The permeability of bound doxorubicin is assumed to follow the same relation.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
116
Figure 6-4. Permeability of free doxorubicin as a function of temperature. Curve fitting parameters z1=0.07 and z2=-2.43.
Experimental data extracted from [144].
(2) Liposome Encapsulated Doxorubicin
For a baseline temperature of 34oC, extracellular concentrations of TSL doxorubicin were found
to have increased by 76-fold and 38-fold upon heating to 45˚C and 42˚C, respectively [145].
Based on these experimental data [145], a relationship between fold increase in permeability and
temperature can be obtained by curve fitting, and the best fitting curve is given by Equation (6-4)
and shown in Figure 6-5.
210120 pp mTT21p
mTT11p0pp ememmm
(6-4)
where T and T0 represent temporal temperature and body temperature, respectively, and the
other terms in Equation (6-4) are curve fitting parameters.
Figure 6-5. Fold increase in permeability of liposome-mediated doxorubicin as a function of temperature. Curve
fitting parameters: mp0=-10.54, To=34, mp11=5.76, mp21=5.78, mp12=5.44, and mp22=5.46. Experimental data of [145]
are adopted.
36 37 38 39 40 41 42 43 44 45 46
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Logar
ithm
ic V
alue
of
Per
mea
bil
ty (
×10
-7m
/s)
Temperature (oC)
Experimental Data
Fitting Curve
33 34 35 36 37 38 39 40 41 42 43 44 45 46
0
10
20
30
40
50
60
70
80
90
Fold
Incr
ease
in P
erm
eabil
ity
Temperature (oC)
Experimental Data
Fitting Curve
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
117
Diffusivity
According to the Stokes-Einstein equation (Equation 3-48), the fold increase in diffusivity as a
function of temperature is
1
*
2
2
*
1
2
1
T
TD
D (6-5)
where D and strands for diffusivity, and μ is the viscosity of solvent. Subscripts 1 and 2
correspond to conditions 1 and 2, respectively. Owing to the lack of relevant data, viscosity
values in Equation (6-5) are assumed to be those of water [146], whose dependence on
temperature is given by Equation (6-6), with the best fitting result shown in Figure 6-6.
2
321exp TmTmmm dddd (6-6)
where T is temporal temperature, and the other terms are parameters used in curve fitting.
Figure 6-6. Viscosity of water as a function of temperature. Curve fitting parameter md1=5.10 md2=-0.03
md3=1.04×10-4
Experimental data extracted from [146]
Rate of Transmembrane Transport
The transmembrane parameter was determined by El-Kareh and Secomb [55] by curve fitting to
data obtained from in vitro experiments [101]. It has been suggested that increased cellular
uptake of doxorubicin with heating will lead to improved outcome when the drug is
administrated simultaneously with hyperthermia [55]. Based on data in [147], there is a 2.2-fold
increase at 42˚C. Here, the fold increase at temperature T is obtained by linear interpolation.
28 32 36 40 44 48 52 56 60 64
40
45
50
55
60
65
70
75
80
85
Wat
er V
isco
city
(10
-5P
as)
Temperature (oC)
Experimental Data
Fitting Curve
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
118
88.724.0 Tktv (6-7)
Perfusion rate
Blood perfusion rate (w) also depends on temperature [141] by the following relation:
DSww 0 (6-8)
where w0 represents the time dependent perfusion at 37˚C, and DS is the degree of vascular stasis
with a value between 0 and 1.
ttRT
E
f deADS0
*exp (6-9)
where T* is the absolute temperature as a function of time t, R is the universal gas constant, and
Af and ΔE represent the frequency factor and the activation energy, respectively.
Plasma Pharmacokinetics of Encapsulated Doxorubicin
After intravenous administration, liposome-encapsulated doxorubicin begins to circulate to all
over the body in circulatory system. Because of the continuous clearance in blood stream and
drug release triggered by environmental temperature, the intravascular concentration of
encapsulated doxorubicin (Clp) is governed by [13]:
lprellplp
lpCkCCL
dt
dC (6-10)
where CLlp and krel refer to the clearance rate and drug release rate, respectively. The initial
concentration of encapsulated doxorubicin is 0.0191 kg/m3, corresponding to a total dose of 50
mg/m2 in literature [62].
Drug Release Rate from Liposome Encapsulation
TSL is designed to release its contents rapidly upon heating [65]. The exact release rate varies
according to the composition of liposome, its preparation procedure and heating temperature [69].
The relation between percentage release and exposure time is found to follow the first-order
kinetics expressed as [148]:
tk
creleRtR
1% (6-11)
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
119
where %R(t) is the percentage of drug released at exposure time t, Rc is the total percentage of
drug released at a given heating temperature. This equation is used to fit experimental data
obtained in the range of 37~42˚C [143], and curve fitting for drug release at 37 o
C and 42oC is
shown in Figure 6-7.
Figure 6-7. Drug release at (a) 37oC and (b) 42
oC. Experimental data extracted from [143].
Note that the TSL formulation used in [143] still allows some drug to be slowly released at 37 oC,
although release at 42oC is much faster. From the best fitting results, release rates at different
temperatures in the range of 37~42˚C are summarized in Table 6-3. Linear interpolation is
performed to obtain release rate at temperatures between the listed temperature points, and the
release rate is assumed to be constant when temperature exceeds 42˚C.
Table 6-3. Release rates at various temperatures
T (oC) 37 38 39 40 41 42
krel (s-1
) 0.00417 0.00545 0.01492 0.02815 0.04250 0.05409
6.2.1.3. Numerical Methods
The mathematical models are implemented into a finite-volume based computational fluid
dynamics (CFD) code, ANSYS FLUENT. The User Defined Scalar (UDS) function is used to
code bioheat transfer model, mass transfer equations governing the drug transport and
pharmacodynamics. The momentum, drug transport and heat transfer equations are discretised
by using the 2nd order UPWIND scheme, and the SIMPLEC algorithm is employed for pressure-
velocity coupling. The convergence is controlled by setting residual tolerances of the momentum
0 200 400 600 800 1000 1200 1400 1600 1800 2000
0
5
10
15
20
25
30
35
40
Rel
ease
Per
centa
ge
(%)
Time (s)
Experimental Data
Fitting Curve
(a)
0 50 100 150 200 250 300 350
0
20
40
60
80
100
120
Rel
ease
Per
centa
ge
(%)
Time (s)
Experimental Data
Fitting Curve
(b)
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
120
equation, drug transport equations and bioheat transfer equations to be 1×10-5, 1×10-8 and 1×10-
10, respectively. As shown in Figure 6-8, interstitial fluid equations are solved first to obtain a
steady-state solution in the entire computational domain. The obtained pressure and velocity are
then introduced at time zero for the simulation of bioheat transfer, drug transport and cellular
uptake. The 2nd order implicit backward Euler scheme is employed to perform the temporal
discretization. A fixed time step size of 0.1 second is chosen after time-step sensitivity tests.
Solving continuity and momentum
equations to obtain steady-state
solution of interstitial fluid field
IFP
IFV
Solving mass
transfer equation
Solving acoustic equation to
generate the acoustic pressure
field in the whole domain
acoustic pressure
Solving bio-heat
transfer model
coupled
Temperature profiles
Drug concentration
Tumour Cell Density
Figure 6-8. Numerical Procedure
Given the time scale of modelling is much shorter than that of tumour growth, all the boundaries
of the tumour and normal tissue are assumed to be fixed. The interface between tumour and
normal tissue is treated as an internal boundary, where all variables are continuous. The relative
pressure and temperature at the surface of normal tissue is specified to be 0 Pa and 37oC, with
zero flux of drug.
6.2.2. Results and Discussion
A coupled mathematical model is used to examine the interplays among the various steps
involved in the delivery of TSL-mediated chemotherapy. Basic features of the interstitial fluid
field are the same as those discussed in Chapters 4 and 5, which will not be repeated here.
Results presented below are focused on spatial and temporal distributions of temperature and
drug concentration in blood, tumour and normal tissue.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
121
6.2.2.1. Temperature Profile
For clinical application, HIFU heating for 30~60 minutes after administration of TSL are
recommended [142]. In an experimental study [149], 10 minutes HIFU heating which was
interleaved with a 5 minutes cooling down period was repeated 3 times. Therefore, simulations
have been carried out for TSL-mediated drug delivery with simultaneously HIFU heating for
durations of 10 minutes, 30 minutes and 60 minutes, respectively. Results are compared with a
control case without heating. The two focused regions presented in Figure 6-3 are heated
sequentially during the treatment, and the sonication duration for each region is 1 second, which
is chosen based on previous studies.
The time course of temperature with 10 minutes heating is shown in Figure 6-9. Variations of
maximum and spatial-mean temperature in blood, tumour and normal tissues have a similar
pattern in response to the heating schedule. The maximum temperature in tumour reaches the
target value of 42oC rapidly, and is then maintained at this level owing to the temperature control
mode governed by Equation (6-1) until the heating period ends. This is followed by a fall in
temperature during the cooling down stage. The lower temperature in blood is because the
continuous perfusion of blood is effective in removing heat from the focus region, and this
forced convection via blood is the main mechanism of heat transfer in tumour and normal tissues.
Results show that temperatures fall back to body temperature within 20 minutes after the
cessation of heating.
0 5 10 15 20 25 30
37
38
39
40
41
42
43
Max
imum
Tem
per
ature
(oC
)
Time (min)
Tumour
Normal Tissue
Blood in the Entire Domain
(a)
0 5 10 15 20 25 30
37.0
37.5
38.0
38.5
39.0
39.5
40.0
40.5
Spat
ial
Mea
n T
emper
ature
(oC
)
Time (min)
Tumour
Nomral Tissue
Blood in Tumour
Blood in Normal Tissue
(b)
Figure 6-9. Variations of temperature in tumour, blood and normal tissue as a function of time for 10 minutes
heating. (a) Maximum and (b) spatial-mean temperature.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
122
Since the focus regions for direct heating are located inside the tumour (shown in Figure 6-3),
temperature in tumour increases faster and is higher than in normal tissues, whereas, temperature
elevation in normal tissues depends on heat transfer from tumour, which in turn causes heat loss
hence reduction in tumour temperature. However, since heat is continuously carried away by the
circulating blood, blood temperature in the tumour region is always lower than tumour
temperature. The large difference in spatial-mean temperature between tumour and normal
tissues (shown in Figure 6-9 (b)) indicates that the release of doxorubicin mainly occurs in the
tumour region, thereby achieving localised treatment and reducing the risk of side effects. Since
blood in the tumour region is directly heated by HIFU, its mean temperature is higher than that in
the surrounding normal tissue.
Spatial distributions of temperature in tissues and blood at 10 minutes after heating are shown in
Figure 6-10. It is clear that the highest temperature is located in the focus regions where tumour
and blood are directly heated by HIFU. Temperature elevation in the surrounding regions is a
consequence of convective and conductive heat transfer from the focus regions, and temperature
in normal tissues is mostly around 37oC, lower than in the tumour. This temperature distribution
can successfully enhance doxorubicin release in tumour.
41.6
41.0
40.3
39.7
39.0
38.3
37.6
37.0
T (oC)
38.6
38.3
38.1
37.9
37.7
37.5
37.2
37.0
T (oC)
Tissue Temperature Blood Temperature
Figure 6-10. Spatial distribution of temperature in (a) tissue and (b) blood at 10min after heating starts.
Figure 6-10 (a) also shows that tissue temperature varies gradually from tumour to normal tissue,
owing to similar thermal properties of tumour and normal tissues summarized in Table 6-1. The
temperature distribution in blood shown in Figure 6-10 (b) has a similar pattern to tissue
temperature but with a narrower range, so that the intravascular release of doxorubicin is much
higher in HIFU focus regions than in the periphery.
Temporal variations of temperature for 30 and 60 minutes heating are shown in Figure 6-11.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
123
Identical shapes can be found in each compartment for a given heating duration. The spatial
mean temperature of tumour tissue remains at 40oC, which is lower than the most effective
release temperature.
0 10 20 30 40 50 60
37
38
39
40
41
42
43
Max
imum
Tem
per
ature
(oC
)
Time (min)
Tumour
Normal Tissue
Blood in the Entire Domain
(a)
0 10 20 30 40 50 60
37.0
37.5
38.0
38.5
39.0
39.5
40.0
40.5
Spat
ial
Mea
n T
emper
ature
(oC
)
Time (min)
Tumour
Normal Tissue
Blood in Tumour
Blood in Normal Tissue
(b)
0 10 20 30 40 50 60 70 80 90
37
38
39
40
41
42
43
44
Tumour
Normal Tissue
Blood in the Entire Domain
Max
imum
Tem
per
ature
(oC
)
Time (min)
(c)
0 10 20 30 40 50 60 70 80 90
37.0
37.5
38.0
38.5
39.0
39.5
40.0
40.5
41.0S
pat
ial
Mea
n T
emper
ature
(oC
)
Time (min)
Tumour
Normal Tissue
Blood in Tumour
Blood in Normal Tissue
(d)
Figure 6-11. Variations of temperature in tumour, blood and normal tissue as a function of time for various heating
durations. (a) Maximum and (b) spatial-mean temperature for 30 minutes heating; (c) Maximum and (d) spatial-
mean temperature for 60 minutes heating.
6.2.2.2. Drug Concentration
Doxorubicin is initially encapsulated in TSL which are administrated intravenously and carried
by the blood stream into the tumour and its surrounding normal tissue. The release of
doxorubicin from TSL is triggered by elevated temperature upon HIFU heating. Therefore, the
concentration of encapsulated and released doxorubicin in blood should be examined when
evaluating its anticancer efficacy. The time courses of encapsulated doxorubicin intravascular
concentration with various heating durations are compared in Figure 6-12. Because of the short
administration period which can be ignored in comparison with the heating duration, drug
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
124
concentration achieves its peak at the very beginning of treatment. Owing to rapid clearance in
the circulatory system and increased release as temperature rises, the intravascular concentration
of encapsulated doxorubicin falls continuously until it drops to 0.
0 5 10 15 20 25 30
0.0
5.0x10-3
1.0x10-2
1.5x10-2
2.0x10-2
Enca
psu
late
d D
oxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (min)
No-Heating
10 min Heating
30 min Heating
60 min Heating
9 10 11 12 13 14 15 16 17 18
0.0
1.0x10-4
2.0x10-4
3.0x10-4
4.0x10-4
5.0x10-4
6.0x10-4
7.0x10-4
Enca
psu
late
d D
oxoru
bic
in I
ntr
avas
cula
r
Conce
ntr
atio
n i
n T
um
our
(kg/m
3)
Time (min)
(a)
0 5 10 15 20 25 30
0.0
5.0x10-3
1.0x10-2
1.5x10-2
2.0x10-2
No-Heating
10 min Heating
30 min Heating
60 min Heating
Enca
psu
late
d D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (min)
9 10 11 12 13 14 15 16 17 18
0.0
3.0x10-4
6.0x10-4
9.0x10-4
1.2x10-3
1.5x10-3
1.8x10-3
Enca
psu
late
d D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (min)
(b)
Figure 6-12. Spatial-mean concentration of encapsulated doxorubicin in blood stream as a function of time (dose =
50 mg/m2). (a) in tumour, and (b) in normal tissue.
Compared with the control case without HIFU heating, encapsulated doxorubicin concentration
in tumour (Figure 6-12 (a)) reduces faster upon HIFU heating, owing to enhanced release rate in
response to temperature raise. It appears that encapsulated doxorubicin concentration drops to
almost zero after about 20 minutes, and the duration of heating (10, 30 and 60 mins) has little
effect on encapsulated doxorubicin concentration in tumour. Since normal tissue temperature is
approximately 2~3 degrees lower than tumour temperature during HIFU heating, the effect of
heating on encapsulated doxorubicin concentration in normal tissue is trivial (Figure 6-12 (b)).
1.23×10-3
1.00×10-3
0.78×10-3
0.55×10-3
0.33×10-3
0.10×10-3
C (kg/m3)
Figure 6-13. Encapsulated doxorubicin concentration in blood at 10 min after heating
The spatial distribution of encapsulated doxorubicin concentration in blood is shown in Figure
6-13 at 10 minutes after heating; it clearly demonstrates that there is little encapsulated
doxorubicin left in and around HIFU focus regions.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
125
The time courses of encapsulated doxorubicin concentration in tumour extracellular space are
presented in Figure 6-14 (a) for various heating durations. Compared with the case without
heating, encapsulated doxorubicin accumulates faster in the extracellular space of the tumour as
soon as HIFU heating begins. This is attributed to enhanced vascular permeability in response to
temperature elevation. However, the peak concentration is lower and occurs earlier in treatments
with hyperthermia. The effect of the heating duration can only be noticed after heating ends.
0 5 10 15 20 25 30
0.0
2.0x10-5
4.0x10-5
6.0x10-5
8.0x10-5
1.0x10-4
1.2x10-4
No-Heating
10 min Heating
30 min Heating
60 min Heating
Enca
psu
late
d D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (min)
(a)
0 5 10 15 20 25 30
-5.0x10-4
0.0
5.0x10-4
1.0x10-3
1.5x10-3
2.0x10-3
2.5x10-3
3.0x10-3
Enca
psu
late
d D
oxoru
bic
in F
lux (
kg/m
3s)
Time (min)
Net Gain
Source: Transvascular Exchange
Sink: Drug Release
(b)
27.0 27.5 28.0 28.5 29.0 29.5 30.0-2.0x10
-7
0.0
2.0x10-7
4.0x10-7
Enca
psu
late
d D
oxoru
bic
in F
lux (
kg/m
3s)
Time (hr)
Figure 6-14. Spatial-mean concentration (a) and flux (b) of encapsulated doxorubicin in tumour extracellular space
as a function of time.
Encapsulated doxorubicin concentration in tumour extracellular space depends on the balance
between the source term which accounts for drug transport from blood vessels, and the sink term
which is related to drug release from TSL. The time courses of each term and net grain in the 60-
minute heating period are presented in Figure 6-14 (b). The source term reaches its peak very
rapidly and then falls off as the intravascular concentration reduces. As a result of increased
extracellular concentration and enhanced release rate, the sink term increases initially until it
catches up with the source term at around 2 minutes after treatment begins, leading to zero
followed by negative net gain. As a consequence, encapsulated doxorubicin extracellular
concentration starts to fall, which in turn causes the sink term to decrease and finally reaching
zero concentration.
Similar patterns can be found in normal tissues as shown in Figure 6-15. Quantitative
comparison shows that encapsulated doxorubicin concentration in normal tissues is 3 orders of
magnitude lower than that in tumour, indicating that TSL-mediated drug delivery could
effectively avoid liposome accumulation in normal tissue.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
126
0 5 10 15 20 25 30 35 40 45 50 55 60
0.0
1.0x10-8
2.0x10-8
3.0x10-8
4.0x10-8
5.0x10-8
6.0x10-8
7.0x10-8
No-Heating
10 min Heating
30 min Heating
60 min Heating
Enca
psu
late
d D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (min)
Figure 6-15. Spatial-mean concentration of encapsulated doxorubicin in extracellular space of normal tissue as a
function of time.
The spatial concentration of encapsulated doxorubicin in extracellular space at different time
points is presented in Figure 6-16. Since blood vessel walls in tumour are leakier with larger
pores than those in normal tissue [110, 150], TSL particles are designed in a size range that
allows them to pass through the vessel walls in tumour only, thereby achieving effective
accumulation in the extracellular space. As shown in Figure 6-16, encapsulated doxorubicin
mainly accumulates in tumour, and drugs in normal tissue (shown in Figure 6-15) come from
tumour by migration in the extracellular space.
1.03×10-4
8.83×10-5
7.36×10-5
5.88×10-5
4.42×10-5
2.95×10-5
1.48×10-5
5.00×10-8
C (kg/m3)
1 min 2 min 5 min
1.03×10-4
8.83×10-5
7.36×10-5
5.88×10-5
4.42×10-5
2.95×10-5
1.48×10-5
5.00×10-8
C (kg/m3)
10 min 15 min 30 min
Figure 6-16. Spatial distribution of encapsulated doxorubicin extracellular concentration in tumour and surrounding
normal tissues at various time points with 60 minutes heating.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
127
Encapsulated doxorubicin concentration in the tumour is highly heterogeneous upon HIFU
heating. Tumour centre has lower concentration than in periphery with time proceeds, owing to
the fast release of doxorubicin enhanced by HIFU heating in and around the focus regions.
Because of the reduction of encapsulated doxorubicin concentration in the blood stream, the
corresponding extracellular concentration decreases in the entire tumour and finally becomes to
uniform at 30 minutes when the temperature falls back to 37 oC.
0.0 0.5 1.0 1.5 2.0 2.5 3.0
0.0
5.0x10-4
1.0x10-3
1.5x10-3
2.0x10-3
2.5x10-3
3.0x10-3
3.5x10-3
4.0x10-3
No-Heating
10 min Heating
30 min Heating
60 min Heating
Fre
e D
oxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
2.6 2.7 2.8 2.9 3.02.4x10
-5
2.8x10-5
3.2x10-5
3.6x10-5
4.0x10-5
Fre
e D
oxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
(a)
0.0 0.5 1.0 1.5 2.0 2.5 3.0
0.0
2.0x10-3
4.0x10-3
6.0x10-3
8.0x10-3
1.0x10-2
1.2x10-2
No-Heating
10 min Heating
30 min Heating
60 min HeatingB
ound D
oxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
(b)
2.5 2.6 2.7 2.8 2.9 3.07.0x10
-5
8.0x10-5
9.0x10-5
1.0x10-4
1.1x10-4
1.2x10-4
1.3x10-4
Bound D
oxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
0.0
5.0x10-4
1.0x10-3
1.5x10-3
2.0x10-3
2.5x10-3
3.0x10-3
3.5x10-3
No-Heating
10 min Heating
30 min Heating
60 min Heating
Fre
e D
oxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
(c)
3.5 3.6 3.7 3.8 3.9 4.02.0x10
-6
2.5x10-6
3.0x10-6
3.5x10-6
4.0x10-6
4.5x10-6
5.0x10-6
5.5x10-6
Fre
e D
oxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
0.0
2.0x10-3
4.0x10-3
6.0x10-3
8.0x10-3
1.0x10-2
1.2x10-2
No-Heating
10 min Heating
30 min Heating
60 min Heating
Bound D
oxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
3.5 3.6 3.7 3.8 3.9 4.06.0x10
-6
9.0x10-6
1.2x10-5
1.5x10-5
1.8x10-5
Bound D
oxoru
bic
in I
ntr
avas
cula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
(d)
Figure 6-17. Doxorubicin extracellular concentration with various heating duration as a function of time. (a) free and
(b) doxorubicin concentration in tumour, and (c) free and (d) doxorubicin concentration in normal tissue.
The time courses of free and bound doxorubicin intravascular concentration in tumour and its
surrounding normal tissue are shown in Figure 6-17. Doxorubicin is released from TSL upon
heating, leading to an increase in intravascular concentration of drug in free and bound forms.
However, drug concentrations fall after reaching their peaks owing to the reduction of
encapsulated doxorubicin in blood. The predicted results show that hyperthermia is capable of
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
128
enhancing drug release and thereby increasing the concentration in blood in the tumour region,
whereas heating makes little difference to intravascular concentration in normal tissue.
0 1 2 3 4 5 6
0.0
5.0x10-4
1.0x10-3
1.5x10-3
2.0x10-3
2.5x10-3
3.0x10-3
3.5x10-3
No-Heating
10 min Heating
30 min Heating
60 min Heating
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
5.5 5.6 5.7 5.8 5.9 6.0
6.0x10-6
9.0x10-6
1.2x10-5
1.5x10-5
1.8x10-5
2.1x10-5
2.4x10-5
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in T
um
our
(kg/m
3)
Time (hr)
(a)
0 1 2 3 4 5 6
0.0
2.0x10-3
4.0x10-3
6.0x10-3
8.0x10-3
1.0x10-2
No-Heating
10 min Heating
30 min Heating
60 min Heating
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
5.5 5.6 5.7 5.8 5.9 6.02.0x10
-5
3.0x10-5
4.0x10-5
5.0x10-5
6.0x10-5
7.0x10-5
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
(b)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
0.0
4.0x10-4
8.0x10-4
1.2x10-3
1.6x10-3
2.0x10-3
No-Heating
10 min Heating
30 min Heating
60 min Heating
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
3.5 3.6 3.7 3.8 3.9 4.06.0x10
-6
9.0x10-6
1.2x10-5
1.5x10-5
1.8x10-5
2.1x10-5
Fre
e D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
(c)
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
0.0
1.0x10-3
2.0x10-3
3.0x10-3
4.0x10-3
5.0x10-3
6.0x10-3
No-Heating
10 min Heating
30 min Heating
60 min Heating
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
3.5 3.6 3.7 3.8 3.9 4.02.0x10
-5
3.0x10-5
4.0x10-5
5.0x10-5
6.0x10-5
Bound D
oxoru
bic
in E
xtr
acel
lula
r C
once
ntr
atio
n
in N
orm
al T
issu
e (k
g/m
3)
Time (hr)
(d)
Figure 6-18. Doxorubicin extracellular concentration with various heating duration as a function of time. (a) free and
(b) doxorubicin concentration in tumour, and (c) free and (d) doxorubicin concentration in normal tissue
Temporal profiles of free and bound doxorubicin concentration in tumour extracellular space and
its surrounding normal tissue are compared in Figure 6-18 between HIFU heating and no-heating.
Regardless of the heating duration, both free and bound doxorubicin concentrations increase
rapidly during the initial period following drug administration. Doxorubicin concentrations in
tumour interstitial space increase faster and reach a much higher peak upon heating. Although
doxorubicin concentration drops to a low level as time proceeds, longer heating duration is able
to maintain a slightly higher concentration. Comparing doxorubicin extracellular concentrations
in Figures 6-18 with the concentration in blood in Figure 6-17, the concentration curves have
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
129
identical shapes for a given heating duration. This means that drug concentration in blood has a
direct influence on the extracellular concentration for both free and bound doxorubicin.
Comparisons show that doxorubicin concentration in both free and bound forms in tumour is
increased by up to 11.26 % upon heating, owing to enhanced drug release, whereas the influence
of HIFU heating on doxorubicin concentration in normal tissue is trivial. On the one hand,
encapsulated doxorubicin concentration in extracellular space of normal tissue is limited (shown
in Figure 6-15 and 6-16), hence less free doxorubicin that can be released. On the other hand,
although there is a large amount of encapsulated doxorubicin in blood, drug release in most
regions of normal tissue is not enhanced as the temperature is close to 37oC.
Spatial distributions of doxorubicin intratumoural concentration and survival fraction of tumour
cells at 10 minutes after administration are displayed in Figure 6-19. It is clear that doxorubicin
concentration and survival tumour cells are non-uniform, which is not surprising since the
tumour and drug transport properties are enhanced by temperature rise, and as such, the position
of HIFU focus regions is expected to have a strong influence on the spatial distribution of drug
and cell killing.
3.39×10-3
3.36×10-3
3.34×10-3
3.31×10-3
3.29×10-3
3.26×10-3
C (kg/m3)
3.20×10-3
3.15×10-3
3.10×10-3
3.05×10-3
3.00×10-3
2.95×10-3
C (kg/m3)
2.68
2.49
2.30
2.12
1.93
1.75
C (ng/105cells)
Free Doxorubicin Intravascular
Concentration
Free Doxorubicin Extracellular
Concentration
Intracellular Concentration in
Tumour
1.02×10-2
1.01×10-2
1.00×10-2
9.93×10-3
9.84×10-3
9.76×10-3
C (kg/m3)
9.60×10-3
9.44×10-3
9.28×10-3
9.12×10-3
8.96×10-3
8.80×10-3
C (kg/m3)
99.43
99.41
99.38
99.36
99.34
99.32
f (%)
Bound Doxorubicin Intravascular
Concentration
Bound Doxorubicin Extracellular
Concentration
Survival Fraction of
Tumour Cells
Figure 6-19. Spatial distribution of doxorubicin concentration and cell survival fraction in tumour at 10 minutes.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
130
Intracellular concentration is the highest in tumour centre where HIFU focus regions are located,
and it decreases radially towards tumour periphery with the lowest level at the right and bottom
corners. As a consequence, cell killing is most effective in the tumour centre, while the cytotoxic
effect is insufficient at the right and bottom corners which are the furthest from the HIFU focus
regions. Therefore, the heterogeneous enhancement presented in Figure 6-19 demonstrate that
TSL-mediated drug delivery with HIFU heating is capable of improving treatment efficacy in
designated tumour regions, and careful selection of focus regions may help achieve desired
therapeutic effect in the entire tumour.
Temporal profiles of intracellular concentration with various heating durations are presented in
Figure 6-20. The intracellular concentration displays a sharp rise at the beginning until it reaches
a peak, and then decreases. Compared with the case with no-heating, the rate of increase in
intracellular concentration improves upon heating. Increasing heating duration results in higher
peaks and higher intracellular concentration levels over time.
0 1 2 3 4 5 6 7 8 9
0
1
2
3
4
5
6
7
No-Heating
10 min Heating
30 min Heating
60 min Heating
Intr
acel
lula
r C
once
ntr
atio
n i
n T
um
or
(ng/1
05ce
lls)
Time (hr)
8.5 8.6 8.7 8.8 8.9 9.04.0x10
-3
6.0x10-3
8.0x10-3
1.0x10-2
1.2x10-2
Intr
acel
lula
r C
once
ntr
atio
n i
n T
um
or
(ng/1
05ce
lls)
Time (hr)
Figure 6-20. Doxorubicin intracellular concentration with heating and no-heating as a function of time.
6.2.2.3. Doxorubicin Cytotoxic Effect
The anticancer effectiveness is evaluated by calculating changes in the percentage of tumour
cells using a pharmacodynamics model detailed in Chapter 3. As can be seen from Figure 6-21,
the percentage of tumour cells starts to fall immediately after drug administration. There is a
rapid cell killing phase in the first few hours after heating in all treatments, but the duration of
this phase depends strongly on the heating duration. This is followed by a slower cell killing
phase which lasts for approximately 2 hours (6~8 hours after administration). With further
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
131
reduction in the intracellular concentration, the rate of cell killing and physiological degradation
is overtaken by that of cell proliferation, and as a result, the percentage of tumour cells begins to
increase. Results in Figure 6-20 suggest that anticancer effectiveness can be improved by
modifying the heating duration.
0 2 4 6 8 10 12 14 16 18
70
75
80
85
90
95
100
No-Heating
10 min Heating
30 min Heating
60 min HeatingS
urv
ival
Fra
ctio
n o
f T
um
our
Cel
ls (
%)
Time (hr)
Figure 6-21. Survival cell fraction in treatment with various heating duration as a function of time.
Relative to the control case with no-heating, improvement in the intracellular concentration and
cell killing with various heating durations are compared in Figure 6-22 (a) and (b), respectively.
Results show that improvements in intracellular concentration and cell killing are non-linearly
related to heating duration. This may be attributed to the fast reduction of encapsulated
doxorubicin concentration in blood and the non-linear relationship of drug transport described in
Chapter 3.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
60 min30 min
Impro
vem
ent
in I
ntr
acel
lula
r C
once
ntr
atio
n
(ng/1
05ce
lls)
10 min
(a)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
60 min30 min
Impro
vem
ent
in C
ell
Kil
ling (
%)
10 min
(b)
Figure 6-22. Improvement by HIFU heating. (a) intracellular concentration, and (b) cell killing
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
132
Although simulation results presented so far demonstrate that TSL drug delivery with HIFU
introduced hyperthermia can improve the cytotoxic effect of doxorubicin, the gain in quantitative
terms is quite small when compared to the control case. Possible reasons are given below.
(1) Unstable liposome. The simulated TSL formulation [143] is unstable at body temperature.
Ideally, TSL should be designed to release its contents upon hyperthermia only, with no release
at physiological temperature. However, this is difficult to achieve in vivo. Undesirable release at
37 oC can reduce the intravascular concentration of encapsulated doxorubicin and thereby
resulting in high concentration of free doxorubicin in blood. As a consequence, less doxorubicin
can be released upon hyperthermia. High concentration of doxorubicin in blood and normal
tissue may increase the risk of side effects, such as cardiotoxicity. Therefore, further
improvement in liposome formulation to prevent drug release at body temperature is essential to
future clinical application of TSL-mediated chemotherapy.
(2) Non-optimal focus region. Tumour cell killing is mostly enhanced in HIFU focus regions.
The rate of doxorubicin release is insufficient in the rest of the tumour as the local temperature is
lower than 42 oC, at which drug releases is most effective. Therefore, heating strategy needs to
be optimised in order to generate a uniform temperature profile close to the desired temperature
within tumour; these include careful design of the number and location of focus regions, heating
duration and HIFU scanning method.
6.3. Comparison of 2-D and 3-D Models
All the numerical results on TSL-mediated drug delivery presented so far are based on a 2-D
model reconstructed from a selected MR image of a transverse section. However, a real tumour
and its surrounding normal tissues are 3-D, and the acoustic pressure as well as the resultant
temperature field are highly heterogeneous upon HIFU heating. Hence, it is necessary to evaluate
the need for 3-D simulation by quantitative comparison of results based on 2-D and 3-D models.
For this purpose, the 3-D model shown in Figure 4-23 is adopted with an improved mesh to meet
the requirement of simulation on acoustic profiles upon HIFU heating, and the final mesh
consists of 4,161,964 tetrahedral elements which have been tested to produce grid independent
solutions.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
133
Numerical simulation has been carried out with only one focus region located in the tumour
centre. Comparison of the maximum temperature within the tumour is shown in Figure 6-23. No
difference can be observed between the 2-D and 3-D models owing to the temperature control
mode described in Equation (6-1).
0 1 2 3 4 5 6 7 8 9 10
37
38
39
40
41
42
43
Tem
per
ature
(oC
)
Time (s)
2-D Model
3-D Model
Figure 6-23. Comparison of maximum temperature of tumour in 2-D and 3-D model.
Temporal profiles of spatial-mean temperature of tissues and blood based on the 2-D and 3-D
models are compared in Figure 6-24 and Figure 6-25, respectively. The predicted temperature is
higher in 2-D model. This over-prediction is attributed to the acoustic pressure distribution which
is highly non-uniform. The focus region to tumour size ratio is 0.17 (mm2/mm2) in the 2-D and
0.03 (mm3/mm3) in 3-D models. Since the focus region covers more tumour regions in the 2-D
model, the predicted temperature rises faster than in the 3-D model. Therefore, the predicted
drug release and cell killing are expected to be more effective in 2-D model simulations.
0 1 2 3 4 5 6 7 8 9 10
37.0
37.2
37.4
37.6
37.8
38.0
Tem
per
ature
(oC
)
Time (s)
2-D Model
3-D Model
(a)
0 1 2 3 4 5 6 7 8 9 10
37.00
37.01
37.02
37.03
37.04
(b)
Tem
per
ature
(o
C)
Time (s)
2-D Model
3-D Model
Figure 6-24. Comparison of spatial mean temperature in 2-D and 3-D model. (a) tumour and (b) normal temperature.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
134
0 1 2 3 4 5 6 7 8 9 10
37.00
37.05
37.10
37.15
37.20
(a)
Tem
per
atur
(oC
)
Time (s)
2-D Model
3-D Model
0 1 2 3 4 5 6 7 8 9 10
37.000
37.002
37.004
37.006
37.008
37.010
Tem
per
ature
(oC
)
Time (s)
2-D Model
3-D Model
(b)
Figure 6-25. Comparison of spatial mean temperature in 2-D and 3-D model. (a) blood temperature in tumour, and
(b) blood temperature in normal tissue.
By incorporating bioheat transfer and drug transport in a realistic tumour model, the presents
study has provided insights into TSL-mediated drug delivery to solid tumours. Nevertheless, the
mathematical model employed here involves a number of assumptions.
First, physical complexities including cavitation and the non-linear progression of ultrasound are
ignored. It has been shown that the effects of these complexities are minor for a local intensity in
the range of 100~1000 W/cm2 and the peak negative pressure in the range of 1~4 MPa [151,
152]. Given that the simulated transducer parameters (acoustic intensity is 120 W/cm2, peak
negative pressure is 1.98 MPa) are within the specified ranges, these physical complexities are
not considered to be important at the conditions examined here.
Second, the influence of temperature elevation on tumour properties is complicated. Hence not
all variations, such as temperature-dependent change in plasma clearance, are included in the
current model. Detailed descriptions and mathematical models are required in the future studies
in order to examine these effects.
Third, focus regions are ideally positioned in the tumour, without considering movement
trajectory of ultrasound transducer, vascular distribution and other micro-environment in tumour.
Moreover, only 3 heating durations have been studied. In order to have a more comprehensive
investigation, complex heating strategies should be simulated. These include release mode,
duration/timing of heating and various release rates.
6. THERMO-SENSITIVE LIPOSOME-MEDIATED DRGU DELIVERY TO SOLID TUMOUR
135
6.4. Summary
In this chapter, thermo-sensitive liposome-mediated drug delivery with various heating durations
are compared based on a 2-D prostate tumour model with a realistic geometry. The results
indicate that thermo-sensitive liposome-mediated drug delivery is a possible method to achieve
localised treatment in cancer therapy. Heating duration is a controllable parameter to improve the
efficacy of anticancer treatment.
7. Conclusion
In this chapter, main conclusions and limitations of the present study are summarized.
This is followed by suggestions for further works in anticancer drug delivery
research.
7. CONCLUSION
137
7.1. Main Conclusion
In this thesis, a mathematical model comprising drug transport and bioheat transfer is presented
to simulate the delivery of anticancer drugs, either in their free form or encapsulated in thermo-
sensitive liposomes. Anticancer efficacy of the delivered drugs is predicted based on a
pharmacodynamics model. Model parameters describing physiological and biological properties
of tissue and drugs are derived from experimental data in the literature. The model has been
applied to realistic tumours reconstructed from MRI in order to (a) understand the transport steps
of non-encapsulated drugs (Chapter 4), (b) elucidate the effects of tumour properties (Chapter 5),
and (c) examine the coupled effect of heat transfer and drug transports under thermo-sensitive
liposome-mediated delivery (Chapter 6).
The main findings are as follows:
(1) Better anticancer efficacy is achieved with administration methods that are able to maintain a
high level of extracellular concentration over the entire treatment period.
In direct delivery of non-encapsulated drugs, continuous infusion is much more effective than
bolus injection in improving the cytotoxic effect of doxorubicin on tumour cells and reducing
peak levels of doxorubicin in blood stream. Although rapid continuous infusion with a high
peak intracellular concentration leads to faster cell killing at the beginning of the treatment,
slow infusion modes give better overall treatment efficacy.
Increasing drug dose can increase concentration levels in the entire tumour and thereby
improving doxorubicin cytotoxicity. Whilst consideration must be given to the limitation on
doses used in clinical treatment owing to potential side effects, which are caused by
doxorubicin in the circulatory system and normal tissues, the highest tolerable dose gives the
best therapeutic efficacy.
Multiple-administration can improve the efficacy of anticancer treatments, while significantly
reducing the peak intravascular concentration, thereby lowering the risk of cardiotoxicity
caused by doxorubicin in blood stream. Treatment outcomes may be improved by increasing
the number of administrations. Although there is a slight reduction in predicted anticancer
7. CONCLUSION
138
effectiveness, reducing the total dose while increasing the number of continuous infusions is
effective in reducing drug concentration in blood.
(2) Interstitial fluid pressure is not shown to be sensitive to microvasculature density in the
interior of tumour, implying that drug transport by pressure-induced convection is weak
except in a thin layer close to the interface of tumour/normal tissue. However, better
anticancer effectiveness is achieved in tumour regions with relatively low but non-zero
microvasculature density, owing to reduced loss of drugs back to the circulatory system.
(3) Spatial-mean transvascular transport is strong in small tumour which is well-vascularised,
owing to the low spatial-mean interstitial fluid pressure and dense microvasculature.
However, anticancer effectiveness is compromised in small tumour as a result of enhanced
reverse diffusion of drug to the blood circulation.
(4) Thermo-sensitive liposome-mediated drug delivery with hyperthermia induced by high
intensity focused ultrasound is effective in releasing encapsulated anticancer drugs within
targeted regions, thereby achieving localised treatment. Heating duration is an important
factor in determining the efficacy of anticancer treatment. Increasing heating duration to 1
hours is beneficial in help maintain higher concentration of anticancer drugs and improve
predicted treatment efficacy.
(5) Under the assumption of homogeneous and uniform tissue and drug transport properties, 2-D
models of tumours without explicit representation of microvasculature provide comparable
results to corresponding 3-D models as far as spatial-averaged parameters are concerned.
Nevertheless, 3-D model is preferred for comprehensive details about spatial distributions of
drug concentration and cell killing, especially when heterogeneous microvasculature density
and heat transfer upon HIFU heating are considered.
7.2. Limitations
The mathematical models described in Chapter 3 involve a number of assumptions which relate
to descriptions of tumour properties, drug transport and HIFU heating. Details of these are given
below.
7. CONCLUSION
139
7.2.1. Tumour Property
In the present study, tumour properties except for cell density are considered to be time-
independent. In reality, tumour growth is a dynamic process which requires variations of tumour
properties with time. The effect of this assumption is considered to be minor for simulations with
a time window of a few hours only. However, variations of tumour properties should be
considered when simulating treatment over a very long time period.
Except for the tumour microvasculature density studied in Chapter 5, tumour properties are
assumed to be homogeneous. However, these properties are likely to be heterogeneous and non-
uniform, which will affect the transport of anticancer drugs to tumour cells, and eventually the
effectiveness of an anticancer treatment.
The present study can be regarded as a qualitative investigation for predictions of trends since all
model parameters, apart from tumour geometry, correspond to average and representative values
extracted from the literature. More information is required for type-specific and patient-specific
studies.
7.2.2. Drug Transport Model
Tumour vascular network is treated as a source term in the governing equations instead of being
modelled explicitly. As a consequence, the effect of intravascular transport and geometric
features of tumour vasculature are not included. While this assumption may be an over-
simplification, it is practical given the lack of information on realistic tumour vasculature
geometry and the complexity in direct modelling of blood flow in a vascular network coupled
with interstitial drug transport.
Furthermore, only the effect of intratumoural microenvironment on drug transport is considered.
However, anticancer drugs may in turn alter the intratumoural microenvironment. For example,
drugs such as doxorubicin also attack blood vessels [138] by reducing microvasculature density
in tumour. This effect can be incorporated by using a more detailed description with a two-way
interaction.
Drug efflux from cells depends on the transmembrane protein, P-Glycoprotein (P-gp), which is
capable of effectively reducing intracellular concentration of structurally unrelated cytotoxic
7. CONCLUSION
140
agents [153], such as doxorubicin [154] and cisplatin [155]. In the present study, this efflux is
included without considering variations of P-gp concentration over time [99]. This limitation
can be overcome by employing an improved transmembrane model.
7.2.3. HIFU Heating Model
The proposed HIFU heating model is a first step towards developing a comprehensive numerical
platform for thermo-sensitive liposome-mediated drug delivery system. The employed transducer
only consists of one element. However, transducers with a few hundreds of elements are required
and have been adopted in clinical use [156]. Moreover, sonication is controlled by a binary
feedback loop with a modified algorithm [157] based on the adjustment of the ultrasound phase
generated by each element.
7.3. Suggestions for Future Work
The long term goal of computational investigations of drug delivery to solid tumour is to develop
an experiment-assisted computational modelling framework spanning across multiple spatial and
time-scales to advance the understanding of drug transport process and provide guidance for
improvement in anticancer treatments. Therefore, several possible modelling tasks have been
identified and these are described below to give an overall idea about how to expand the
modelling work described in this thesis.
7.3.1. Realistic Blood Vessel Geometry
The morphological feature of blood vessels in tumour is significantly different from those in
normal tissue, leading to highly abnormal organization and connection between vessels [20, 21].
Experimental [158] and numerical [102] studies showed that not all the tumour cells could be
exposed to sufficient drugs during treatment. Real micro-vasculature networks reconstructed
from advanced imaging techniques may help to examine the effect of blood vessel geometric
features on drug delivery and the resultant cytotoxic effect. Modelling of blood flow can then be
coupled with intravascular drug transport. Such a model can provide spatial-temporal profiles of
drug concentration and cytotoxic effect to gain more insights into the effect of tumour-specific
blood vessel network on anticancer effectiveness in different types of tumours.
7. CONCLUSION
141
7.3.2. Nutrient Transport
Tumour growth needs nutrient which is also carried by the circulating blood. Highly proliferating
tumour cells and insufficient supply of nutrient, such as oxygen and glucose, may result in
hypoxia in localised tumour regions, and cause tumour cell arrest. Given that many anticancer
drugs (e.g. doxorubicin) act on proliferating cells because their working mechanism is to stop
cell duplication by interacting with DNA [3], treatment efficacy may be compromised in regions
with insufficient supply of oxygen and other nutrients. On the other hand, arrested tumour cells
can re-proliferate after nutrition in the surrounding environment is restored, or tumour cells may
die from nutrition shortage, resulting in the formation of necrotic core. It is important to point out
that blood supply is closely related to the distribution of functional microvasculature. Therefore,
it would make the model more realistic by incorporating nutrient transport, especially when
heterogeneous distribution of microvasculature is considered.
7.3.3. Model Validation
The present study serves to provide predictions of qualitative trends, hence detailed model
validations are not included except for the HIFU acoustic model. This limitation is attributed to
both the numerical model and experiments. On the one hand, not all model parameters are
available for a specified tumour, especially when considering the unique condition of each
patient; on the other hand, important results, such as intracellular concentration and cell density,
cannot be directly measured in vivo. Animal experiments may help to provide comparable data
for validation. Advanced optical methods, such as high-performance liquid chromatography, can
potentially be used to obtain in vivo data for validations in the future.
7.3.4. Type / Patient-Specific Tumour
The properties of tumour and normal tissue vary with tissue type, growth stage and individual
conditions of a patient [41, 45]. These variations may influence drug transport processes and
eventually the anticancer efficacy. Numerical study can be used to predict these differences and
provide suggestions in order to optimize treatments. However, detailed type-specific and patient-
specific information is required.
7. CONCLUSION
142
7.3.5. Chemotherapy Drugs
Since different anticancer drugs can be combined in clinical use [3, 159, 160], it would be useful
to simulate drug transport and cell killing for different combinations of drugs. Most of
chemotherapy drugs experience similar transport processes in delivery to tumour cells, including:
transport within/through blood vessels, migration in interstitial space and cell uptake. Besides
doxorubicin, the basic mathematical model presented here can be applied to various drugs, such
as 5-fluorouracid, BCNU, cisplatin, Methotrexate and antibody (e.g. Fab and IgG). Necessary
modifications should be made in terms of drug-specific properties and behaviour. For instance,
5-fluorouracid and cisplatin do not easily bind with proteins [17].
7.3.6. HIFU Heating Strategy
Heating strategy is essential in thermo-sensitive liposome-mediated drug delivery systems.
Further computational studies should be carried out to design an optimal heating strategy that
aims to generate a uniform temperature distribution within designated tumour regions in order to
improve local treatment efficacy while reducing side effects to normal organs. HIFU scanning
method of multiple-element transducer can be designed to maintain tumour temperature above
the phase transition temperature of liposome while minimising heating of untargeted regions.
Moreover, the predicted anticancer efficacy can be influenced by a number of factors which
should be optimised for the treatment of different types of tumours; these include release mode
(intra- and extra-vascular triggered release), duration and timing of heating and release rate of
chemotherapy drugs.
In summary, a mathematical model for drug delivery in free and encapsulated forms has been
developed in order to provide qualitative and mechanistic understanding of anticancer drug
transport and cytotoxicity in solid tumours. The developed numerical platform can serve as a
foundation for further comprehensive investigations on anticancer drug delivery.
143
Appendix: Publications during the Project
Journal papers
[1] Wenbo Zhan, Xiao Yun Xu, 2013, A mathematical model for thermosensitive liposomal
delivery of Doxorubicin to solid tumour., J Drug Deliv, Vol:2013, ISSN:2090-3014
[2] Wenbo Zhan, Wladyslaw Gedroyc, Xiao Yun Xu, 2014, Mathematical Modelling of Drug
Transport and Uptake in a Realistic Model of Solid Tumour. Protein Pept Lett, Vol: 21, Issue
No.11, 2014: 1146-56.
[3] Wenbo Zhan, Wladyslaw Gedroyc, Xiao Yun Xu, 2014. Effect of Heterogeneous
Microvasculature Distribution on Drug Delivery to Solid Tumour. Journal of Physics: D. 47,
2014: 475401
Conference Presentations
[1] Wenbo Zhan, Wladyslaw Gedroyc, Xiao Yun Xu. Computational Study of Drug Transport in
Realistic Models of Solid Tumour. AIChE Annual Conference, Pittsburgh, US, 2012
[2] Wenbo Zhan, Wladyslaw Gedroyc, Xiao Yun Xu. Computational Study of Drug Transport in
Realistic Models of Solid Tumour. Mathways into Cancer II, Carmona, Sevilla, Spain, 2013
[3] Wenbo Zhan, Xiao Yun Xu. Simulation of HIFU Heating in Solid Tumour: Comparison of
Different Temperature Control Modes. 3rd
International Conference on Computational
Methods for Thermal Problems, Lake Bled, Slovenia, 2014
144
Appendix: Nomenclature
Symbols
A1 parameter of pharmacokinetic model for doxorubicin
A2 parameter of pharmacokinetic model for doxorubicin
A3 parameter of pharmacokinetic model for doxorubicin
Af frequency factor
b chord from the ultrasound transducer centre to the radiator boundary
c specific heat
Cbe bound doxorubicin concentration in interstitial fluid
Cbp bound doxorubicin concentration in blood plasma
Cfe Free doxorubicin concentration in the interstitial fluid
Cfp free doxorubicin concentration in blood plasma
CGd tracer concentration in the interstitial fluid
Ci intracellular doxorubicin concentration
CLbp plasma clearance of bound doxorubicin
Cle liposome encapsulated drug concentration in the interstitial fluid
CLfp plasma clearance of free doxorubicin
Cp_Gd tracer concentration in blood
Cv doxorubicin concentration in blood plasma
d distance from a point on the radiator to a field point
Dbe diffusion coefficient of the bound doxorubicin-protein
Dc tumour cell density
Dd drug dose
Dfe diffusion coefficient of free doxorubicin
Dl diffusion coefficient of liposomes
DS degree of vascular stasis
DW diffusion coefficient in water
EC50 drug concentration producing 50% of fmax
Fbe bound doxorubic gain from blood vessels
145
Ffl doxorubicin loss of doxorubicin to the lymphatic vessels per unit volume of tissue
Ffp doxorubicin gained from the blood capillaries in tumour and normal tissues
Ffp free doxorubicin gained from the capillaries in tumour and normal tissues
Fll loss of liposomes through the lymphatic vessels per unit volume of tissue
Flp liposomes gained from the capillaries in tumour and normal tissues
Fly interstitial fluid absorption rate by the lymphatics per unit volume of tumour tissue
fmax cell-kill rate constant
Fv interstitial fluid loss from blood vessels per unit volume of tumour tissue
h depth of the concave surface
Ia ultraound intensity
k thermal conductivity
ka doxorubicin-protein binding rate
kac acoustic velocity potential
kB Boltzmann's constant
kci PI control parameter
kcp PI control parameter
kd doxorubicin-protein dissociation rate
ke michaelis constant for transmembrane transport
kg cell physiological degradation rate
ki michaelis constant for transmembrane transport
Kly hydraulic conductivity of the lymphatic wall
kp cell proliferation rate constant
krel release rate of doxorubicin from liposome
Kv hydraulic conductivity of the microvascular wall
md inceased fold of diffusivity
mp inceased fold of permeability
mtv inceased fold of transmembrane rate
MW molecular weight
P permeability
pa acoustic pressure
Pbe permeability of vessel wall to bound doxorubicin
146
Pef trans-capillary Peclet number
Pfe permeability of vasculature wall to free doxorubicin
pi interstitial fluid pressures
ply intra-lymphatic pressure
pv vascular pressures
q ultrasound power deposition
r radius of spherical particle
R universal gas constant
s percentage of bound doxorubicin
S/V surface area of blood vessels per unit volume of tumour tissue
(S/V)* relative value of surface area of blood vessel per tumour size
Sb net rate of doxorubicin gained from association/dissociation with bound doxorubicin-
protein
SGd MRI signal with tracer
SGd_0 MRI signal without tracer
Si net rate of doxorubicin gained from the surrounding environment
Sl net rate of liposomes gained from the surrounding environment
Slp liposomes gained from plasma
Sly/V surface area of lymphatic vessels per unit volume of tumour tissue
Sr released drug in the interstitial fluid
Su influx/efflux from tumour cells
Sv net rate of doxorubicin gained from the blood/lymphatic vessels
T temperature
T* absolute temperature
T0 body temperature
t time
td infusion duration
Tn normal body temperature
u0 velocity constant
va speed of ultrasound
VB plasma volume
147
Vmax rate of trans-membrane transport
VT tumour volume
W diagonal matrix
w perfusion rate of blood flow
w0 perfusion rate of blood flow at body temperature
ΔE activation energy
Tensor
C matrix of the inertial loss term
F Darcian resistance to fluid flow through porous media
I unit tensor
u normal velocity of a slightly curved radiating surface
v velocity of the interstitial fluid
Greek Letter
α ultrasound absorption coefficient
α1 compartment clearance rate of doxorubicin
α2 compartment clearance rate of doxorubicin
α3 compartment clearance rate of doxorubicin
ɛ cellular efflux function
ζ cellular uptake function
κ permeability of the interstitial space
λ wave length of ultrasound
μ dynamic viscosity of interstitial fluid
πi osmotic pressure of interstitial fluid
πv osmotic pressure of the plasma
ρ density
σ osmotic reflection coefficient
σd osmotic reflection coefficient for the drug molecules
148
σT average osmotic reflection coefficient for plasma protein
τ stress tensor
φ tumour fraction extracellular space
Ψ acoustic velocity potential
ω angular frequency of ultrasound transducer
Abbreviation
BSA body surface area
BSNC bis-chloroethylnitrosourea
CFD computational fluid dynamics
FVM finite volume method
Gd-DTPA gadopentetate dimeglumine
HIFU high intensity focused ultrasound
IFP interstitial fluid pressure
IFV interstitial fluid velocity
MRI magnetic resonance image
MW molecular weight
PEG polyethylene glycol
PDT photodynamic therapy
P-gp P-Glycoprotein
TSL thermo-sensitive liposome
UDS user defined scalar
149
Bibliography
1. King, R.J.B. and M.W. Robins, Cancer biology. 3rd ed. 2006, Harlow: Pearson/Prentice Hall.
xvii, 292 p.
2. Weinberg, R.A., The biology of cancer. 2007, New York: Garland Science. xix, 844 p.
3. Goodman, L.S., et al., Goodman & Gilman's the pharmacological basis of therapeutics. 11th ed.
2005, New York: McGraw-Hill. 2021 p.
4. Kennedy, J.E., High-intensity focused ultrasound in the treatment of solid tumours. Nature
reviews cancer, 2005. 5(4): p. 321-327.
5. Braathen, L.R., et al., Guidelines on the use of photodynamic therapy for nonmelanoma skin
cancer: an international consensus. Journal of the American Academy of Dermatology, 2007.
56(1): p. 125-143.
6. Morgan, R.A., et al., Cancer regression and neurologic toxicity following anti-MAGE-A3 TCR
gene therapy. Journal of immunotherapy (Hagerstown, Md.: 1997), 2013. 36(2): p. 133-51.
7. Sarkar, S., et al., Chemoprevention gene therapy (CGT) of pancreatic cancer using perillyl
alcohol and a novel chimeric serotype cancer terminator virus. Current molecular medicine, 2014.
14(1): p. 125-140.
8. Sclafani, F., et al., TP53 mutational status and cetuximab benefit in rectal cancer: 5-year results
of the EXPERT-C trial. Journal of the National Cancer Institute, 2014. 106(7): p. 121.
9. Aryal, M., et al., Multiple treatments with liposomal doxorubicin and ultrasound-induced
disruption of blood–tumor and blood–brain barriers improve outcomes in a rat glioma model.
Journal of Controlled Release, 2013. 169(1): p. 103-111.
10. Allen, T.M. and P.R. Cullis, Liposomal drug delivery systems: from concept to clinical
applications. Advanced drug delivery reviews, 2013. 65(1): p. 36-48.
11. Bangham, A. and R. Horne, Negative staining of phospholipids and their structural modification
by surface-active agents as observed in the electron microscope. Journal of molecular biology,
1964. 8(5): p. 660-IN10.
12. Zhang, X., et al., Towards an understanding of the release behavior of temperature-sensitive
liposomes: a possible explanation of the “pseudoequilibrium” release behavior at the phase
transition temperature. Journal of liposome research, 2013. 23(3): p. 167-173.
13. Gasselhuber, A., et al., Targeted drug delivery by high intensity focused ultrasound mediated
hyperthermia combined with temperature-sensitive liposomes: computational modelling and
preliminary in vivovalidation. Int J Hyperthermia, 2012. 28(4): p. 337-48.
150
14. Au, J.L., et al., Determinants of drug delivery and transport to solid tumors. J Control Release,
2001. 74(1-3): p. 31-46.
15. Jain, R.K., Transport of molecules in the tumor interstitium: a review. Cancer Res, 1987. 47(12):
p. 3039-51.
16. Skinner, S.A., G.M. Frydman, and P.E. O'Brien, Microvascular structure of benign and
malignant tumors of the colon in humans. Dig Dis Sci, 1995. 40(2): p. 373-84.
17. Jang, S.H., et al., Drug delivery and transport to solid tumors. Pharm Res, 2003. 20(9): p. 1337-
50.
18. Clark, E.R. and E.L. Clark, Microscopic observations on the growth of blood capillaries in the
living mammal. American journal of anatomy, 1939. 64(2): p. 251-301.
19. Cliff, W.J., Kinetics of Wound Healing in Rabbit Ear Chambers, a Time Lapse Cinemicroscopic
Study. Q J Exp Physiol Cogn Med Sci, 1965. 50: p. 79-89.
20. Dreher, M.R., et al., Tumor vascular permeability, accumulation, and penetration of
macromolecular drug carriers. Journal of the National Cancer Institute, 2006. 98(5): p. 335-344.
21. Less, J.R., et al., Microvascular architecture in a mammary carcinoma: branching patterns and
vessel dimensions. Cancer Res, 1991. 51(1): p. 265-73.
22. Less, J.R., et al., Geometric resistance and microvascular network architecture of human
colorectal carcinoma. Microcirculation, 1997. 4(1): p. 25-33.
23. Skinner, S.A., P.J. Tutton, and P.E. O'Brien, Microvascular architecture of experimental colon
tumors in the rat. Cancer Res, 1990. 50(8): p. 2411-7.
24. Yamaura, H. and H. Sato, Quantitative studies on the developing vascular system of rat hepatoma.
J Natl Cancer Inst, 1974. 53(5): p. 1229-40.
25. Wu, N.Z., et al., Increased microvascular permeability contributes to preferential accumulation
of Stealth liposomes in tumor tissue. Cancer Res, 1993. 53(16): p. 3765-70.
26. Yuan, F., et al., Microvascular permeability and interstitial penetration of sterically stabilized
(stealth) liposomes in a human tumor xenograft. Cancer Res, 1994. 54(13): p. 3352-6.
27. Chang, Y.S., et al., Mosaic blood vessels in tumors: frequency of cancer cells in contact with
flowing blood. Proceedings of the National Academy of Sciences, 2000. 97(26): p. 14608-14613.
28. di Tomaso, E., et al., Mosaic tumor vessels: cellular basis and ultrastructure of focal regions
lacking endothelial cell markers. Cancer research, 2005. 65(13): p. 5740-5749.
29. McDonald, D.M. and P.L. Choyke, Imaging of angiogenesis: from microscope to clinic. Nat Med,
2003. 9(6): p. 713-25.
151
30. Heuser, L.S. and F.N. Miller, Differential macromolecular leakage from the vasculature of
tumors. Cancer, 1986. 57(3): p. 461-4.
31. Winkler, F., et al., Kinetics of vascular normalization by VEGFR2 blockade governs brain tumor
response to radiation: role of oxygenation, angiopoietin-1, and matrix metalloproteinases.
Cancer Cell, 2004. 6(6): p. 553-63.
32. Hobbs, S.K., et al., Regulation of transport pathways in tumor vessels: role of tumor type and
microenvironment. Proc Natl Acad Sci U S A, 1998. 95(8): p. 4607-12.
33. Yuan, F., et al., Vascular permeability and microcirculation of gliomas and mammary
carcinomas transplanted in rat and mouse cranial windows. Cancer Res, 1994. 54(17): p. 4564-8.
34. Bundgaard, M., Transport pathways in capillaries-in search of pores. Annual review of
physiology, 1980. 42(1): p. 325-336.
35. Firrell, J.C., G.P. Lewis, and L.J. Youlten, Vascular permeability to macromolecules in rabbit
paw and skeletal muscle: a lymphatic study with a mathematical interpretation of transport
processes. Microvasc Res, 1982. 23(3): p. 294-310.
36. Simionescu, N., M. Simionescu, and G.E. Palade, Open junctions in the endothelium of the
postcapillary venules of the diaphragm. J Cell Biol, 1978. 79(1): p. 27-44.
37. Lum, H. and A.B. Malik, Regulation of vascular endothelial barrier function. Am J Physiol, 1994.
267(3 Pt 1): p. L223-41.
38. Seymour, L.W., Passive tumor targeting of soluble macromolecules and drug conjugates. Critical
reviews in therapeutic drug carrier systems, 1991. 9(2): p. 135-187.
39. Endrich, B., et al., Tissue perfusion inhomogeneity during early tumor growth in rats. J Natl
Cancer Inst, 1979. 62(2): p. 387-95.
40. Jain, R.K., Transport of molecules across tumor vasculature. Cancer and Metastasis Reviews,
1987. 6(4): p. 559-593.
41. Rubin, P. and G. Casarett, Microcirculation of tumors part I: anatomy, function, and necrosis.
Clinical radiology, 1966. 17(3): p. 220-229.
42. Jain, R.K., Determinants of tumor blood flow: a review. Cancer Res, 1988. 48(10): p. 2641-58.
43. Peters, W., et al., Microcirculatory studies in rat mammary carcinoma. I. Transparent chamber
method, development of microvasculature, and pressures in tumor vessels. Journal of the National
Cancer Institute, 1980. 65(3): p. 631-642.
44. Hori, K., et al., Characterization of heterogeneous distribution of tumor blood flow in the rat.
Cancer Science, 1991. 82(1): p. 109-117.
152
45. Vaupel, P., F. Kallinowski, and P. Okunieff, Blood flow, oxygen and nutrient supply, and
metabolic microenvironment of human tumors: a review. Cancer research, 1989. 49(23): p. 6449-
6465.
46. Saltzman, W.M., Drug delivery: engineering principles for drug therapy. 2001: Oxford
University Press.
47. Maeda, H., et al., Tumor vascular permeability and the EPR effect in macromolecular
therapeutics: a review. J Control Release, 2000. 65(1-2): p. 271-84.
48. Boucher, Y. and R.K. Jain, Microvascular pressure is the principal driving force for interstitial
hypertension in solid tumors: implications for vascular collapse. Cancer research, 1992. 52(18): p.
5110-5114.
49. Baxter, L.T. and R.K. Jain, Transport of fluid and macromolecules in tumors. I. Role of
interstitial pressure and convection. Microvasc Res, 1989. 37(1): p. 77-104.
50. Baxter, L.T. and R.K. Jain, Transport of fluid and macromolecules in tumors. II. Role of
heterogeneous perfusion and lymphatics. Microvasc Res, 1990. 40(2): p. 246-63.
51. Goh, Y.M., H.L. Kong, and C.H. Wang, Simulation of the delivery of doxorubicin to hepatoma.
Pharm Res, 2001. 18(6): p. 761-70.
52. Cusack, B.J., et al., Doxorubicin and doxorubicinol pharmacokinetics and tissue concentrations
following bolus injection and continuous infusion of doxorubicin in the rabbit. Cancer
chemotherapy and pharmacology, 1993. 32(1): p. 53-58.
53. Legha, S.S., et al., Reduction of doxorubicin cardiotoxicity by prolonged continuous intravenous
infusion. Annals of Internal Medicine, 1982. 96(2): p. 133-139.
54. Hortobagyi, G., et al., Decreased cardiac toxicity of doxorubicin administered by continuous
intravenous infusion in combination chemotherapy for metastatic breast carcinoma. Cancer, 1989.
63(1): p. 37-45.
55. El-Kareh, A.W. and T.W. Secomb, A mathematical model for comparison of bolus injection,
continuous infusion, and liposomal delivery of doxorubicin to tumor cells. Neoplasia, 2000. 2(4):
p. 325-38.
56. Gabizon, A.A., Pegylated liposomal doxorubicin: metamorphosis of an old drug into a new form
of chemotherapy. Cancer Invest, 2001. 19(4): p. 424-36.
57. Soundararajan, A., et al., [(186)Re]Liposomal doxorubicin (Doxil): in vitro stability,
pharmacokinetics, imaging and biodistribution in a head and neck squamous cell carcinoma
xenograft model. Nucl Med Biol, 2009. 36(5): p. 515-24.
58. Bangham, A., M.M. Standish, and J. Watkins, Diffusion of univalent ions across the lamellae of
swollen phospholipids. Journal of molecular biology, 1965. 13(1): p. 238-IN27.
153
59. Litzinger, D.C., et al., Effect of liposome size on the circulation time and intraorgan distribution
of amphipathic poly (ethylene glycol)-containing liposomes. Biochimica et Biophysica Acta
(BBA)-Biomembranes, 1994. 1190(1): p. 99-107.
60. Liu, D., A. Mori, and L. Huang, Role of liposome size and RES blockade in controlling
biodistribution and tumor uptake of GM1-containing liposomes. Biochimica et Biophysica Acta
(BBA)-Biomembranes, 1992. 1104(1): p. 95-101.
61. Gabizon, A., H. Shmeeda, and Y. Barenholz, Pharmacokinetics of pegylated liposomal
Doxorubicin: review of animal and human studies. Clin Pharmacokinet, 2003. 42(5): p. 419-36.
62. Gabizon, A., et al., Prolonged circulation time and enhanced accumulation in malignant exudates
of doxorubicin encapsulated in polyethylene-glycol coated liposomes. Cancer Res, 1994. 54(4): p.
987-92.
63. Vaage, J., et al., Tissue distribution and therapeutic effect of intravenous free or encapsulated
liposomal doxorubicin on human prostate carcinoma xenografts. Cancer, 1994. 73(5): p. 1478-
1484.
64. Batist, G., et al., Reduced cardiotoxicity and preserved antitumor efficacy of liposome-
encapsulated doxorubicin and cyclophosphamide compared with conventional doxorubicin and
cyclophosphamide in a randomized, multicenter trial of metastatic breast cancer. Journal of
Clinical Oncology, 2001. 19(5): p. 1444-1454.
65. Zou, Y., et al., Enhanced therapeutic effect against liver W256 carcinosarcoma with
temperature-sensitive liposomal adriamycin administered into the hepatic artery. Cancer Res,
1993. 53(13): p. 3046-51.
66. Gaber, M.H., et al., Thermosensitive sterically stabilized liposomes: formulation and in vitro
studies on mechanism of doxorubicin release by bovine serum and human plasma. Pharm Res,
1995. 12(10): p. 1407-16.
67. Unezaki, S., et al., Enhanced delivery and antitumor activity of doxorubicin using long-
circulating thermosensitive liposomes containing amphipathic polyethylene glycol in combination
with local hyperthermia. Pharm Res, 1994. 11(8): p. 1180-5.
68. Lindner, L.H., et al., Novel temperature-sensitive liposomes with prolonged circulation time. Clin
Cancer Res, 2004. 10(6): p. 2168-78.
69. Tagami, T., M.J. Ernsting, and S.D. Li, Optimization of a novel and improved thermosensitive
liposome formulated with DPPC and a Brij surfactant using a robust in vitro system. J Control
Release, 2011. 154(3): p. 290-7.
70. de Smet, M., et al., Magnetic resonance imaging of high intensity focused ultrasound mediated
drug delivery from temperature-sensitive liposomes: an in vivo proof-of-concept study. J Control
Release, 2011. 150(1): p. 102-10.
154
71. Hynynen, K., MRIgHIFU: A tool for image‐guided therapeutics. Journal of Magnetic Resonance
Imaging, 2011. 34(3): p. 482-493.
72. Staruch, R., R. Chopra, and K. Hynynen, Localised drug release using MRI-controlled focused
ultrasound hyperthermia. International Journal of Hyperthermia, 2010. 27(2): p. 156-171.
73. Staruch, R.M., et al., Enhanced drug delivery in rabbit VX2 tumours using thermosensitive
liposomes and MRI-controlled focused ultrasound hyperthermia. International Journal of
Hyperthermia, 2012. 28(8): p. 776-787.
74. Sheu, T.W.H., et al., On an acoustics–thermal–fluid coupling model for the prediction of
temperature elevation in liver tumor. International Journal of Heat and Mass Transfer, 2011.
54(17): p. 4117-4126.
75. Qiao, S., et al., Effects of different parameters in the fast scanning method for HIFU treatment.
Medical physics, 2012. 39: p. 5795.
76. Lin, W.-L., R.B. Roemer, and K. Hynynen, Theoretical and experimental evaluation of a
temperature controller for scanned focused ultrasound hyperthermia. Medical physics, 1990. 17:
p. 615.
77. Mougenot, C., et al., Three‐dimensional spatial and temporal temperature control with MR
thermometry‐guided focused ultrasound (MRgHIFU). Magnetic Resonance in Medicine, 2009.
61(3): p. 603-614.
78. Wang, C.-H. and J. Li, Three-dimensional simulation of IgG delivery to tumors. Chemical
Engineering Science, 1998. 53(20): p. 3579-3600.
79. Teo, C.S., et al., Transient interstitial fluid flow in brain tumors: Effect on drug delivery.
Chemical Engineering Science, 2005. 60(17): p. 4803-4821.
80. Tzafriri, A.R., et al., Mathematical modeling and optimization of drug delivery from
intratumorally injected microspheres. Clin Cancer Res, 2005. 11(2 Pt 1): p. 826-34.
81. Weinberg, B.D., et al., Model simulation and experimental validation of intratumoral
chemotherapy using multiple polymer implants. Medical & biological engineering & computing,
2008. 46(10): p. 1039-1049.
82. Kuenen, B.C., et al., Dose-finding and pharmacokinetic study of cisplatin, gemcitabine, and
SU5416 in patients with solid tumors. Journal of clinical oncology, 2002. 20(6): p. 1657-1667.
83. Wihlm, J., et al., Pharmacokinetic profile of high-dose doxorubicin administered during a 6 h
intravenous infusion in breast cancer patients]. Bulletin du cancer, 1997. 84(6): p. 603.
84. Stoller, R.G., et al., Use of plasma pharmacokinetics to predict and prevent methotrexate toxicity.
New England Journal of Medicine, 1977. 297(12): p. 630-634.
155
85. Robert, J., et al., Pharmacokinetics of adriamycin in patients with breast cancer: correlation
between pharmacokinetic parameters and clinical short-term response. European Journal of
Cancer and Clinical Oncology, 1982. 18(8): p. 739-745.
86. Greene, R.F., et al., Plasma pharmacokinetics of adriamycin and adriamycinol: implications for
the design of in vitro experiments and treatment protocols. Cancer Res, 1983. 43(7): p. 3417-21.
87. Deen, W.M., Hindered transport of large molecules in liquid‐filled pores. AIChE Journal, 1987.
33(9): p. 1409-1425.
88. Baxter, L.T. and R.K. Jain, Transport of fluid and macromolecules in tumors. III. Role of binding
and metabolism. Microvasc Res, 1991. 41(1): p. 5-23.
89. Baxter, L.T. and R.K. Jain, Transport of fluid and macromolecules in tumors. IV. A microscopic
model of the perivascular distribution. Microvasc Res, 1991. 41(2): p. 252-72.
90. Eikenberry, S., A tumor cord model for doxorubicin delivery and dose optimization in solid
tumors. Theor Biol Med Model, 2009. 6: p. 16.
91. Netti, P.A., et al., Time-dependent behavior of interstitial fluid pressure in solid tumors:
implications for drug delivery. Cancer Research, 1995. 55(22): p. 5451-5458.
92. Erlichman, C. and D. Vidgen, Cytotoxicity of adriamycin in MGH-U1 cells grown as monolayer
cultures, spheroids, and xenografts in immune-deprived mice. Cancer research, 1984. 44(11): p.
5369-5375.
93. Nederman, T. and J. Carlsson, Penetration and binding of vinblastine and 5-fluorouracil in
cellular spheroids. Cancer chemotherapy and pharmacology, 1984. 13(2): p. 131-135.
94. Nederman, T., J. Carlsson, and K. Kuoppa, Penetration of substances into tumour tissue. Model
studies using saccharides, thymidine and thymidine-5'-triphosphate in cellular spheroids. Cancer
chemotherapy and pharmacology, 1987. 22(1): p. 21-25.
95. Nicholson, K., M. Bibby, and R. Phillips, Influence of drug exposure parameters on the activity
of paclitaxel in multicellular spheroids. European Journal of Cancer, 1997. 33(8): p. 1291-1298.
96. West, G.W., R. Weichselbaum, and J.B. Little, Limited penetration of methotrexate into human
osteosarcoma spheroids as a proposed model for solid tumor resistance to adjuvant
chemotherapy. Cancer research, 1980. 40(10): p. 3665-3668.
97. Erlanson, M., E. Daniel-Szolgay, and J. Carlsson, Relations between the penetration, binding and
average concentration of cytostatic drugs in human tumour spheroids. Cancer chemotherapy and
pharmacology, 1992. 29(5): p. 343-353.
98. Dordal, M.S., et al., Flow cytometric assessment of the cellular pharmacokinetics of fluorescent
drugs. Cytometry, 1995. 20(4): p. 307-314.
156
99. Luu, K.T. and J.A. Uchizono, P-glycoprotein induction and tumor cell-kill dynamics in response
to differential doxorubicin dosing strategies: a theoretical pharmacodynamic model.
Pharmaceutical research, 2005. 22(5): p. 710-715.
100. Friedman, M.H., Principles and models of biological transport. 2008: Springer.
101. Kerr, D.J., et al., Comparative intracellular uptake of adriamycin and 4'-deoxydoxorubicin by
non-small cell lung tumor cells in culture and its relationship to cell survival. Biochem
Pharmacol, 1986. 35(16): p. 2817-23.
102. Liu, C., J. Krishnan, and X.Y. Xu, Investigating the effects of ABC transporter-based acquired
drug resistance mechanisms at the cellular and tissue scale. Integrative Biology, 2013.
103. Doroshow, J.H., Anthracyclines and anthracenediones. Cancer Chemotherapy and Biotherapy:
Principles and Practice. 2nd ed. Philadelphia, Pa: Lippincott-Raven, 1996: p. 409-434.
104. Ozawa, S., et al., Cell killing action of cell cycle phase-non-specific antitumor agents is
dependent on concentration--time product. Cancer Chemother Pharmacol, 1988. 21(3): p. 185-90.
105. Eichholtz-Wirth, H., Dependence of the cytostatic effect of adriamycin on drug concenration and
exposure time in vitro. Br J Cancer, 1980. 41(6): p. 886-91.
106. Gewirtz, D., A critical evaluation of the mechanisms of action proposed for the antitumor effects
of the anthracycline antibiotics adriamycin and daunorubicin. Biochemical pharmacology, 1999.
57(7): p. 727-741.
107. El-Kareh, A.W. and T.W. Secomb, A mathematical model for cisplatin cellular
pharmacodynamics. Neoplasia (New York, NY), 2003. 5(2): p. 161.
108. Lankelma, J., et al., Simulation model of doxorubicin activity in islets of human breast cancer
cells. Biochimica et Biophysica Acta (BBA)-General Subjects, 2003. 1622(3): p. 169-178.
109. Eliaz, R.E., et al., Determination and modeling of kinetics of cancer cell killing by doxorubicin
and doxorubicin encapsulated in targeted liposomes. Cancer research, 2004. 64(2): p. 711-718.
110. Yuan, F., et al., Vascular permeability in a human tumor xenograft: molecular size dependence
and cutoff size. Cancer Res, 1995. 55(17): p. 3752-6.
111. Pennes, H.H., Analysis of tissue and arterial blood temperatures in the resting human forearm. J
Appl Physiol, 1948. 1(2): p. 93-122.
112. Curra, F.P., et al., Numerical simulations of heating patterns and tissue temperature response due
to high-intensity focused ultrasound. Ultrasonics, Ferroelectrics and Frequency Control, IEEE
Transactions on, 2000. 47(4): p. 1077-1089.
113. Filonenko, E. and V. Khokhlova, Effect of acoustic nonlinearity on heating of biological tissue by
high-intensity focused ultrasound. Acoustical Physics, 2001. 47(4): p. 468-475.
157
114. Solovchuk, M.A., et al., Simulation study on acoustic streaming and convective cooling in blood
vessels during a high-intensity focused ultrasound thermal ablation. International Journal of Heat
and Mass Transfer, 2012. 55(4): p. 1261-1270.
115. O'Neil, H., Theory of focusing radiators. The Journal of the Acoustical Society of America, 1949.
21: p. 516.
116. Hilmas, D.E. and E.L. Gillette, Morphometric analyses of the microvasculature of tumors during
growth and after x-irradiation. Cancer, 1974. 33(1): p. 103-10.
117. Pappenheimer, J.R., E.M. Renkin, and L.M. Borrero, Filtration, diffusion and molecular sieving
through peripheral capillary membranes. Am. J. Physiol, 1951. 167(13): p. 2578.
118. Nugent, L.J. and R.K. Jain, Extravascular diffusion in normal and neoplastic tissues. Cancer Res,
1984. 44(1): p. 238-44.
119. Wu, N.Z., et al., Measurement of material extravasation in microvascular networks using
fluorescence video-microscopy. Microvasc Res, 1993. 46(2): p. 231-53.
120. Curry, F.E., V.H. Huxley, and R.H. Adamson, Permeability of single capillaries to intermediate-
sized colored solutes. Am J Physiol, 1983. 245(3): p. H495-505.
121. Ribba, B., et al., A mathematical model of Doxorubicin treatment efficacy for non-Hodgkin's
lymphoma: investigation of the current protocol through theoretical modelling results. Bull Math
Biol, 2005. 67(1): p. 79-99.
122. Gerlowski, L.E. and R.K. Jain, Microvascular permeability of normal and neoplastic tissues.
Microvasc Res, 1986. 31(3): p. 288-305.
123. Granath, K.A. and B.E. Kvist, Molecular weight distribution analysis by gel chromatography on
Sephadex. J Chromatogr, 1967. 28(1): p. 69-81.
124. Saltzman, W.M. and M.L. Radomsky, Drugs released from polymers: diffusion and elimination
in brain tissue. Chemical Engineering Science, 1991. 46(10): p. 2429-2444.
125. Swabb, E.A., J. Wei, and P.M. Gullino, Diffusion and convection in normal and neoplastic
tissues. Cancer Res, 1974. 34(10): p. 2814-22.
126. Tong, R.T., et al., Vascular normalization by vascular endothelial growth factor receptor 2
blockade induces a pressure gradient across the vasculature and improves drug penetration in
tumors. Cancer Res, 2004. 64(11): p. 3731-6.
127. Aukland, K. and R.K. Reed, Interstitial-lymphatic mechanisms in the control of extracellular
fluid volume. Physiol Rev, 1993. 73(1): p. 1-78.
128. Wolf, M.B., P.D. Watson, and D.R. Scott, 2nd, Integral-mass balance method for determination
of solvent drag reflection coefficient. Am J Physiol, 1987. 253(1 Pt 2): p. H194-204.
158
129. Saltiel, E. and W. McGuire, Doxorubicin (adriamycin) cardiomyopathy. West J Med, 1983.
139(3): p. 332-41.
130. Singal, P.K. and N. Iliskovic, Doxorubicin-induced cardiomyopathy. New England Journal of
Medicine, 1998. 339(13): p. 900-905.
131. Barth, T.J. and D.C. Jespersen, The design and application of upwind schemes on unstructured
meshes. 1989.
132. Benet, L.Z. and P. Zia-Amirhosseini, Basic principles of pharmacokinetics. Toxicologic
pathology, 1995. 23(2): p. 115-123.
133. Rodvold, K.A., D.A. Rushing, and D.A. Tewksbury, Doxorubicin clearance in the obese. Journal
of Clinical Oncology, 1988. 6(8): p. 1321-1327.
134. Dubois, A.B., Julie; Mentre, France Mathematical expressions of the pharmacokinetic and
pharmacodynamic models implemented in the PFIM software. University Paris Diderot, 2011.
135. Raghunathan, S., D. Evans, and J.L. Sparks, Poroviscoelastic modeling of liver biomechanical
response in unconfined compression. Annals of biomedical engineering, 2010. 38(5): p. 1789-
1800.
136. Zhao, J., H. Salmon, and M. Sarntinoranont, Effect of heterogeneous vasculature on interstitial
transport within a solid tumor. Microvascular research, 2007. 73(3): p. 224-236.
137. Jain, H.V., J.E. Nör, and T.L. Jackson, Modeling the VEGF–Bcl-2–CXCL8 Pathway in
Intratumoral Agiogenesis. Bulletin of mathematical biology, 2008. 70(1): p. 89-117.
138. Schiffelers, R.M., et al., Anti-tumor efficacy of tumor vasculature-targeted liposomal doxorubicin.
Journal of Controlled Release, 2003. 91(1): p. 115-122.
139. Arap, W., R. Pasqualini, and E. Ruoslahti, Cancer treatment by targeted drug delivery to tumor
vasculature in a mouse model. Science, 1998. 279(5349): p. 377-380.
140. Eatock, M., A. Schätzlein, and S. Kaye, Tumour vasculature as a target for anticancer therapy.
Cancer treatment reviews, 2000. 26(3): p. 191-204.
141. Schutt, D.J. and D. Haemmerich, Effects of variation in perfusion rates and of perfusion models
in computational models of radio frequency tumor ablation. Med Phys, 2008. 35(8): p. 3462-70.
142. Grüll, H. and S. Langereis, Hyperthermia-triggered drug delivery from temperature-sensitive
liposomes using MRI-guided high intensity focused ultrasound. Journal of Controlled Release,
2012. 161(2): p. 317-327.
143. Tagami, T., et al., A thermosensitive liposome prepared with a Cu2+
gradient demonstrates
improved pharmacokinetics, drug delivery and antitumor efficacy. Journal of Controlled Release,
2012. 161(1): p. 142-149.
159
144. Dalmark, M. and H.H. Storm, A Fickian diffusion transport process with features of transport
catalysis. Doxorubicin transport in human red blood cells. J Gen Physiol, 1981. 78(4): p. 349-64.
145. Gaber, M.H., et al., Thermosensitive liposomes: extravasation and release of contents in tumor
microvascular networks. Int J Radiat Oncol Biol Phys, 1996. 36(5): p. 1177-87.
146. Keenan, J.H. and F.G. Keyes, Thermodynamic properties of steam : including data for the liquid
and solid phases. 1936, New York, London ;: Wiley ; Chapman & Hall.
147. Nagaoka, S., et al., Intracellular uptake, retention and cytotoxic effect of adriamycin combined
with hyperthermia in vitro. Jpn J Cancer Res, 1986. 77(2): p. 205-11.
148. Afadzi, M., et al. Ultrasound stimulated release of liposomal calcein. in Ultrasonics Symposium
(IUS), 2010 IEEE. 2010. IEEE.
149. Ranjan, A., et al., Image-guided drug delivery with magnetic resonance guided high intensity
focused ultrasound and temperature sensitive liposomes in a rabbit Vx2 tumor model. Journal of
controlled release, 2012. 158(3): p. 487-494.
150. Monsky, W.L., et al., Augmentation of transvascular transport of macromolecules and
nanoparticles in tumors using vascular endothelial growth factor. Cancer Research, 1999. 59(16):
p. 4129-4135.
151. Bailey, M., et al., Physical mechanisms of the therapeutic effect of ultrasound (a review).
Acoustical Physics, 2003. 49(4): p. 369-388.
152. Hariharan, P., M. Myers, and R. Banerjee, HIFU procedures at moderate intensities—effect of
large blood vessels. Physics in medicine and biology, 2007. 52(12): p. 3493.
153. Marbeuf-Gueye, C., et al., Inhibition of the P-glycoprotein and multidrug resistance protein-
mediated efflux of anthracyclines and calceinacetoxymethyl ester by PAK-104P. European
journal of pharmacology, 2000. 391(3): p. 207-216.
154. Maitra, R., et al., Differential effects of mitomycin C and doxorubicin on P-glycoprotein
expression. Biochem. J, 2001. 355: p. 617-624.
155. Demeule, M., M. Brossard, and R. Béliveau, Cisplatin induces renal expression of P-glycoprotein
and canalicular multispecific organic anion transporter. American Journal of Physiology-Renal
Physiology, 1999. 277(6): p. F832-F840.
156. Köhler, M.O., et al., Volumetric HIFU ablation under 3D guidance of rapid MRI thermometry.
Medical physics, 2009. 36(8): p. 3521-3535.
157. Enholm, J.K., et al., Improved volumetric MR-HIFU ablation by robust binary feedback control.
Biomedical Engineering, IEEE Transactions on, 2010. 57(1): p. 103-113.
160
158. Primeau, A.J., et al., The distribution of the anticancer drug Doxorubicin in relation to blood
vessels in solid tumors. Clinical Cancer Research, 2005. 11(24): p. 8782-8788.
159. Murad, A.M., et al., Modified therapy with 5‐fluorouracil, doxorubicin, and methotrexate in
advanced gastric cancer. Cancer, 1993. 72(1): p. 37-41.
160. von der Maase, H., et al., Long-term survival results of a randomized trial comparing
gemcitabine plus cisplatin, with methotrexate, vinblastine, doxorubicin, plus cisplatin in patients
with bladder cancer. Journal of Clinical Oncology, 2005. 23(21): p. 4602-4608.